EMRFD Message Archive 11669

Message Date From Subject
11669 2015-09-25 19:18:47 farhanbox@gmail.c... Atlas 210x reboot

Allison

atlas 210x is about the only commercially produced transceiver that i have wished to own. it was the first transceiver that i ever saw and operated for a few minutes. it has a ladder filter, a diode dbm mixer and a strong post mix amp. here are is my list of things to do with it.

1. change the vfo to use a low noise jfet. add a Huff n Puff stabilizer.

2. replace the band pass filters with better triple tuned ones with low loss 

3. the if board has to go. thr MC1350p was a noisy device. the noise would go up as the gain was reduced. the Haywards design even with two cascode stages will do wonders.

4. reduce the power to 10 watts with one of the RD devices.


i would also want to change the mixer to KISS mixer with 2N7000s and eliminate the post mix amp but that would be changing the soul of the machine.

 

the other trouble was too much tx gain in three stages. but then i operated it only for a few minutes into an inverted V. so i cant comment upon it from experience. how much does an atlas 210 carcass go for these days?

- f

11670 2015-09-25 22:42:04 Kirk Kleinschmidt Re: Atlas 210x reboot
Farhan,

On ePay here in the States, untested "fixer uppers" go for about $100, while used units that apparently work fetch about $200.

Best of luck. Perhaps you'll turn it into a 210-BITX ?    :)

--Kirk, NT0Z
My book, "Stealth Amateur Radio," is now available from
www.stealthamateur.com and on the Amazon Kindle (soon)




11671 2015-09-26 06:55:11 Ashhar Farhan Re: Atlas 210x reboot
hmm .. that's an idea. the trouble with being afar from the americas is that the postage is often costlier than the purchase. as a result, i time these things to coincide with the migratory patterns of my academician friends who come to india during their winter and summer breaks.
i am over-prepared for the 210x :-)
- f

11675 2015-09-26 13:03:48 bob_ledoux Re: Atlas 210x reboot
Be careful for what you wish for.  Ashhar will drive the Atlas 210 price skyward. 

Remember how Bill, N2CQR, drove up the price of Drake 2-B's by his comments on the Soldersmoke podcast.   Heathkit QF-1, Q-multipliers with their vernier variable caps were cheap until Bill let that cat out of the bag.  Now, I'm forced to use varicaps!
11679 2015-09-26 17:27:27 kb1gmx Re: Atlas 210x reboot

Some of this is likely going to get me some hate mail from Atlas lovers 

and owners.  But just for a moment put emotion aside.  I really don't 

hate it.  I am however being critical with the advantage that in 40 years 

we(collective and myself)  learned a lot about radio and transceiver 

design.  Its a groundbreaking radio with things that were a result of 

pushing the bleeding edge.


1. change the vfo to use a low noise jfet. add a Huff n Puff stabilizer.


Been done by others.  Has potential but putting a FET the VFO alone is not a cure.


The problem of the VFO is its band switched and for some bands runs at frequencies

that are over 20mhz, it takes a lot of finesse even with a stabilizer to keep it from 

drifting excessively fast and popping from lock to lock.  There are two sources 

of drift in the VFO the device and the LC.  The device has a varying LC so its

operating point changes from band to band.  The drift from that is improved by

adjusting bias on each and every unit for best stability.  Then the LC system 

has to be temperature compensated.  Of course that more than 40 years has 

elapsed helps none of that. It doesn't fix the excessive level of harmonics

resulting from its design.  It would still need a digital display as the analog 

is mechanically frail plastic from ageing.


Three fixes that have to be applied in my mind.  Insure the VFO power is 

stable replace their with a low dropout regulator.  Change the bias on teh VFO

oscillator to find the zero drive oscillator operating point.  That's recognizes 

that every transistor used for that function had a different Hfe (and generally 

still do).  By doing that on the one I have under 20hz on 40M is doable in a 

stable temperature desktop.  before that it would drift down at 200hz per 

minute and would never settle.  NOTE on this bias change if you went to 

far it would drift the other way and never settle. also the caps used 

must be replaces with good COG parts first.  It still will drift from heat but 

that's the LC.



The VFO is the biggest single problem.  A Si570 system might work well there.

A DDS with low SFDR can work there too (9951).   Either that or a good single band 

VFO with a offset tracking PLL to get in the range for other bands. The 

challenge is space and keeping it clean.


Did I mention its a tiny radio for then and maybe even now.


2. replace the band pass filters with better triple tuned ones with low loss 


Triple tuned with low loss, somewhat contradictory.  That and space, your

mod  would entail be making your own boards.  Making new boards however 

can be an improvement alone by going two sided and better grounding.

Maybe shielding improvements.


In some places and cases better parts alone would be a net improvement.


3. the if board has to go. thr MC1350p was a noisy device. the noise would go up as the gain was reduced. the Haywards design even with two cascode stages will do wonders.


This is less a problem, the reason for that is by time the AGC starts your dealing 

with fairly large signals and added noise is not an issue.   The radio has other 

compromises that have far more impact but yes there are a few 1350s.  Haywood

is not wrong, only that in this radio its impact is low.


The radio is unique as its a low gain SSB (very unique for then) and the receiver

gain is limited to the post mixer amp and 1 stage of 1350 and a DBM product 

detector.  For the receiver that all of the RF gain used and about 15-20db of 

that is offsetting the input filter, 1st DBM, Crystal filter, and 2nd DBMs losses.  

Its usefulness on the air is enough audio gain and a low noise front end without 

a preamp (so high loss input filters can kill you there).  The liability is the Rx input 

filters are also the TX band pass filters and that would impact TX gain distribution.  


Also going from 1350 to a Hycas mean a total rework of the ALC and AGC system.

The 1350 has a threshold of about 5V and give positive from there, Hycas has a 

threshold about 2/3rds the power rail and goes negative from there.


Change that and your making a new SWR, AGC, and  metering board. 


4. reduce the power to 10 watts with one of the RD devices.


The driver is in the watts range already.  RD parts at 10W are not a absolute panacea.

A good bipolar is a valid choice too.  If anything I'd not use one for that stage but go 

push pull for better power efficiency with RD or BJT.  It would be easy to do that.

The amp itself (driver and finals) bolt onto the back easy to remove and restyle.


The above 4 mods if done would impact most every board in the radio, some would 

have to be a complete overhaul.  It would make an interesting radio.


To understand the radio and its topology read about the "Lichen" in EMRFD

its the same topology and the discussion over stage reuse, gain distribution,

and its liabilities is eyeopening.  The design was high performs to a point and 

that point is limited by the switching CIO and VFO sources and other 

limits inherited.  That's not to say its bad only that you can't make the next 

super radio of it.  Its a good exercise in even with DDS or NCO for the LO and 

modern hardware your get smaller but physics of electronics limit your much 

better it can be.   Sorta once you open the can your get the full distribution 

of worms.


I have a bit more contact with them over the years fixing a few.  I have one 

in my hands as I write but its a nice unmolested one and I'm not ready to 

do terrible things to it.  I'm looking for one with blown finals and "issues" 

as a test bed for ideas.   As I said it would be an interesting radio to modify.


There are other older radios out there for modding as well.  There are a few 

IcKenYae and older TenTec radios that are sold state up the driver and final 

that could be solid stated (new driver and final all solid sate) dumping the 

HV power needs and heat issues like VFO drift.


Plenty of fine radio candidates.



Allison


11680 2015-09-27 03:55:04 Ashhar Farhan Re: Atlas 210x reboot
hmm .. i agree allison, the truth me told i used one almost 25 years ago for a few hours. i have never peeked inside one. most of the new radios confuse the hell out of me. there are so many knobs to fiddle that i am nervous that i'd leave some (like transmit offset or something) turned on. as a result i like simple radios like the Atlas and some TenTecs. 

- f

11681 2015-09-27 04:24:45 blumu Re: Atlas 210x reboot
I agree with both of you. My favourites are my Atlas 210X and Century22 (in addition to
homebrew receivers) - not least because I can understand and maintain them :-)

Michael
UK

----- Original Message -----

11682 2015-09-27 05:21:29 kb1ckt Re: Atlas 210x reboot
I have a relatively simple TenTec, and have managed to leave RIT on just the same. It happens.

shawn kb1ckt

Sent from my iPad

11683 2015-09-27 09:49:51 kb1gmx Re: Atlas 210x reboot
I can see that I tend to fall in that realm.  I seriously not a fan or "menu" radios.  
The FT817 its tiny I understand that  but there are others that just make me scream.

The Atlas is a simple radio and that has appeal to me.  It part of the reason I like my 
Tentec Triton (M540), simple analog radio that works and sounds good.

Allison

11684 2015-09-27 09:57:45 kb1gmx Re: Atlas 210x reboot
Oddly I have four radios I call modern.

Tentec 6n2 my main radio for 6 and 2M  with a pair of amps its the ticket.
Tentec Eagle Heck of a good HF+6..  I have yet to hook it to a computer, may never.
Tentec Triton M540 the analog side rule dial killer receiver for a solid state radio built in 1978.
FT817, small and does DC to daylight (ok HF through 432 skipping 222) its the portable do all.

The rest are HB SSB radios at least count 7 of them all monobander or purpose built and 
another in the works.  They have sparse panels as a  rule.  Volume, tuning, maybe RF gain
or attenuator switch.  One even has a mic gain, had to fill a existing hole with something.

simple is good.  ;)

Allison


11692 2015-09-28 03:52:17 Fernando Krouwel Re: Atlas 210x reboot
Hi Allison, good morning:

I have the 5.52MHz xtal filter (and also the small board containing the BFO/carrier oscillator) for the Atlas transceiver, which I purchased some time ago in a flea market. I would like to use them in a small project, probably a BITX or something like.

Do you have any information about this filter, like input/output impedances and if it is (or not) bidirectional?

73' s
Fernando - PY2FZU



11693 2015-09-28 04:07:38 kb1gmx Re: Atlas 210x reboot
Do you have any information about this filter, like input/output impedance and if it is (or not) bidirectional?

Ok this is from Clint not personal testing.

Filter about 800 ohms.   Its assumed both sides of it are the same.  It would not be hard to 
test that.

All crystal filters are bidirectional for signal.

Find a copy of the manual on line I'd suggest the Atlas 180 as its representative and also simplest.
It would not be hard to duplicate the radio save may be the final amp and the VFO system which 
are the two you would least want to do.  VFO a high end DDS should do and final for low power
there are published options to a few watts.  Oddly enough the whole radio has no parts that I 
know of that are not still available.  Examples are the MC1350 IF amp, LM380 audio, Ca3089
transistor array that can be done with either discrete transistors(3904) or Ca3046 (same part, 
better specs).  About the only thing you cant get new is the finals, filter, and sheet metal.


Allison
11694 2015-09-28 04:24:56 Fernando Krouwel Re: Atlas 210x reboot
Thank you very much!

Fernando

11698 2015-09-28 23:54:04 John Lawson Re: Atlas 210x reboot
What is the physical measurements of the filter, length, width, etc. also does the filter have Network Sciences as the manufacture stamped on the top of the filter case? Any other markings stamped on the filter would be of help.....finally do you know if it is the SSB filter or CW filter.....If so, I may be able to provide you with some information. Let me know as I do have some of their filters in the junk box........John K5IRK

Sent from my iPad

11699 2015-09-29 04:03:57 Fernando Krouwel Re: Atlas 210x reboot
Hi John, good morning:

For more details I would have to look for it, but I remember, yes, it is made by Network Sciences, 5.52MHz and the writings appeared to be simply stamped on its metalic case. It was not with a beautiful looking, so I polished it with a... metal polisher (stamped infos disasappeared then), that removed the last layer of plating, leaving it with a copper looking (before this polishment it had a very ugly looking that was even difficult to read the maker and other infos). One thing that I observed is that it didn' t have the input/output id. Also I was not capable to locate these pins ID in the Atlas Manual.

About the BFO/Carrier oscillator, I tested it and works OK, both sidebands.

I thought about building a simple SSB transceiver or the Progressive Receiver with it.
In our present days, the Progressive Receiver has became easier to build, since now we have Si570, Si5351 and some simple and cheap DDS, wich enable us to build it multiband with single conversion, avoiding those expensive and hard to find (custom made) xtals needed for the converter.

73' s
Fernando - PY2FZU


11700 2015-09-29 04:34:38 John Lawson Re: Atlas 210x reboot
Hi Fernando, give me a day or so to find them in my junk box. Also if you have the schematic diagram which shows the filters, if you can scan that part and send it to me off line at jmlcs2000@yahoo.com it would help me to see if mine are the same. Finally do the physical measurements of the size of the filter case too. I have several of their filters and there is one that is smaller than the others......Finally I'm curious as to the frequencies of the BFO crystals.73, John

Sent from my iPad

11701 2015-09-29 07:21:52 kb1gmx Re: Atlas 210x reboot
With those filters the input and output are not marked and are reversible.

The actual frequency of the carrier osc has to be set (trimmed) in the system 
for best results and is rarely exactly the marked crystal can frequency.


Allison
11702 2015-09-29 08:43:22 atlantaswl Re: Atlas 210x reboot
Hi Fernando,
Concerning your thrid paragraph, I just completed the digital infrastructure for a 16-band progressive communications (W7ZOI and K5IRK) receiver supporting 4 crystal filters.  I am currently using 2 AD9851 for both the VFO and heterodyne oscillator.  All the band pass filters are built and aligned; the IF amp is the hybrid cascode (W7ZOI and WA7MLH). The design is still single conversion for 80 meters, dual conversion for all bands above 80 meters.

Communication to the band-pass filter relay board, the crystal filter relay board, and the two DDS boards is SPI, and parallel to the display.  This has been a fun project. 

73s,

Fred  KK4VEP

11705 2015-09-29 16:33:32 Fernando Krouwel Re: Atlas 210x reboot
Hi Fred,

All I have to say is:
Wow, congratulations!!!

Fernando - PY2FZU


11706 2015-09-30 09:47:11 kb1gmx Re: Atlas 210x reboot
Hi Fred,

A curious question:  Back when the progressive RX was introduced the primary issue in any radio
was a LOs system that was stable.  That meant a 160 or 80M radio.  Even in the article it noted that
multiple conversion had cost in dynamic range by the payback then was a stable receiver and ability 
to later add bands.   Earlier influences were when using IF int he 455 or 500khz range meant multiple conversion to avoid image response above 80M.  But by then a filter in the 9mhz range meant few if 
issues up to 6M and beyond.

Fast forward 40 years:

The image issue is long gone.  We routinely build good IF filters at HF and higher with 5 to 12mhz 
being most common for home brewers and 4.9152Mhz being plenty high enough for 10M.

The VFO issue is solved.  We have SI570 NCO and high end (low SFDR) DDS available the 
AD9851 being an example.  What makes these device appealing is they are easier to implement 
than the multi-loop PLLs commercial radios opted for.

The question: why multiple conversion as single conversion has become easier?  Just dial the 
LO to where you want to be and switch in a new front end filter?

After all everything after the input to the first mixer doesn't change.

As a VHF-UHF  sort there is logic for converters /transverters/ for bands 2M and up.
Up there band switching becomes impractical and dynamic range less an issue but 
noise figure is.

Mostly thinking out loud as someone that have spanned the days when a filter was 
a costly and rare thing (McCoy, SI, and Collins) and you built with what you had to 
now when you make a filter to suit needs.

Allison

11707 2015-09-30 12:18:11 iq_rx Re: Atlas 210x reboot
Hi Allison,

Thank you for giving us the benefit of your thinking out loud.  Something you didn't mention is that single conversion receivers have many fewer internally generated spurious responses.  It is generally possible in narrow band applications with good shielding to build receivers that have no internal birdies across the desired tuning range.  This is critically important when digging for weak signals, either while tuning across the band noise and listening, or using a DSP engine to dig deep.

In the late 1980s I experienced many hours of frustration while tuning for 1296.1 MHz CW signals from 450 miles away, hearing something very weak, tuning back and forth to see if I could discern any keying, and finally turning off the converter and discovering it was still there on the Collins 75S3C IF receiver...an internal birdie.  My first phasing receiver system was designed to replace (and outperform) the Collins 75S3C in that application.  In critical UHF weak signal applications I've used single conversion systems ever since.

An original motivations for all the work I did on phasing receivers and transmitters was that I could build single conversion systems with a VHF or UHF IF, and still get rid of the opposite sideband noise and signals on the other side of zero beat.

My single conversion homebrew stations for HF through 5.7 GHz are all spur-free over the entire range of the tuning dial, something I couldn't say about any of my multi-conversion commercial gear.

That early motivation for my direct conversion receiver experiments sometimes resulted in a humorous disconnect with the QRP community.  My QRP friends all seemed to be on a mission for simplicity and minimum parts count, and I was designing something that would outperform a 75S3C.  Most of the world thinks of direct conversion as a way to reduce cost and complexity.  I've always approached it as a way to improve performance.

Best Regards,

Rick kk7b
11708 2015-09-30 12:53:31 Tayloe, Dan (Noki... Re: Atlas 210x reboot

Amen to non-simple and high performance.  I like to use R/C active filters in my DC phased receivers.  14 poles of low Q active R/C filtering is a joy to listen to, but is not particularly low parts count.  I think I almost killed Dough Hendricks when Norcal kitted the NC2030.  Over 400 parts for a “simple” DC CW transceiver?  Lots and lots of Rs and Cs with too many unique values.  I could do much better today minimizing the number of unique parts, but for a phasing DC receiver, parts = performance.

 

-          Dan, N7VE

 

11709 2015-09-30 20:12:18 kb1gmx Re: Atlas 210x reboot
I learned the multiconversion lesson by building one, ok I'm dense, several.  Why several.
Because I was young, uneducated, and didn't understand spurs.  It taught me well.  I 
learned shielding, circuit bypassing, isolation, and filtering and then to just not go there.

Regardless of the technique some older designs persist for wrong reasons.
That's not saying the progressive was wrong, it was very good.  Only that 
little details were not spelled out as to why and what.   For example If you 
build the progressive and put each converter in its own box and plug it it 
it works well if the main RX is in a box.  I knew someone who knew more 
than me and he put each converter next to each other visible to the main 
chassis.  Havoc, the only word for the result.  He would take the whole 
thing apart box each section and each converter, in the end it worked 
very well.  However he did it twice.  

There is a difference between simple to me the R2 is simple and minimum 
art where even bypass caps are often left out to make a lower parts count.
The later is also known ans Munsoning after "mad man Munson" of early 
TV manufacturing.  He would hover over engineers and start pulling parts 
out till it would stop working and then say put that one back and ship it. 
Simple to me is systems devoid of issues by doing things at audio rather 
than at 10mhz for the same effect.  Its simpler to build and troubleshoot.  
As someone that has designed and built computer systems with many 
hundreds of chips more parts is not complex its just more parts.   A 
lot of simple stages is simple.  Complex is when there are a lot of 
things and any go wrong it all fails and make catastrophically.  
Complex is a test setup that takes an amount of gear to make a lab 
engineer envious to make it work. 

If there is a bottom line and your duplicating a older design read the 
whole article.  Read the other peoples war stories.Look at similar things 
and see what that writer/builder had to say.  Look at the pictures
and take a shot at why the writer/builder did it that way or shouldn't have.
Make a point of building as experimental so if a circuit doesn't work 
as planned it can be replaced or fixed.  Then enjoy the build.


Allison

11712 2015-10-01 08:45:56 atlantaswl Re: Atlas 210x reboot
Hi Allison,
Thanks for responding to my post.  There are a number of reasons for the approach I took.

First, this was the first receiver I've built from the ground up after Kitchen's regenerative
receiver.  The article by Wes and John looked doable with my skill set (mathematics vs
electrical engineering background), but with lots of kit-building experience.

I started with the assumption I would be able to acquire all those crystals for the heterodyne
oscillators.  After some research, I realized I would have to go with digital synthesis.  With 
the dual conversion, the VFO range would be limited to 12.5 to 13.2 MHz for all bands,
except 80 meters.  I am able to turn off the heterodyne oscillator when tuning 80 meters.  
To lessen the effects of any spurs, I could use a low pass filter that will attenuate those
spurs outside this VFO range. Another benefit was the ease of programming the microprocessor 
to control the VFO, set the heterodyne oscillator, apply the offset per band to display the correct 
frequency being received, and select the band pass filter, all with one button push on the front
panel.  I am much more familiar with software engineering than I am with RF design. I am
learning a lot on this forum.

I do plan on quite a bit of shielding.  All band pass filters (14) , the BPF relay board and 80 meter
preselector are on the underside of the chassis.  The VFO and heterodyne oscillator are in
Hammond 1590A enclosures with sma connectors for output and bypass caps for control and
Vcc.  The BFO oscillator (crystal-controlled) will also be shielded similarly on the underside of
the chassis.  The IF amp, both mixers, microprocessor board, digital signal routing board,
crystal filters, relay board for these filters and audio amp are all 'above ground' and are currently 
not in enclosures. I plan on providing enough space on the top side of the chassis between
boards to provide enclosures for any of the boards, if I need it.  All RF signal routing will be
through RG-174.

I've got a lot of SMA male connectors to attach to the various boards.  Does anyone have a
decent SMA torque wrench they could recommend?

73s,
Fred KK4VEP


11713 2015-10-01 09:06:22 Jim Strohm Re: Atlas 210x reboot
Yeah, I've got one that says "5/16" on one end and "Snap-on" in the middle of the handle part.  It's all nice and shiny with triple chrome plating.


While it's not best practice, you'll get "good enough" connections at HF simply by getting SMA connectors somewhat tighter than finger tight with a small wrench.  If you are a stickler, get an inch-pound torque wrench, set it for the manufacturer's specs (you DO know who made your SMAs, right?) for the connectors, and then find a solid way to hook your 5/16" open-end wrench to it.

Then measure by feel how much pressure you need to meet the recommended spec.

Or -- just buy an SMA torque wrench.  I think they're about a hundred bucks new.  I know a guy who occasionally has some for $25 each, but they're surplus, and hence they aren't in current cal.

73
Jim N6OTQ



11714 2015-10-01 11:29:19 Steve Dick Re: Atlas 210x reboot
Hi Dan. I’ve been a fan of the NC2030 but it was no longer available when I first came across it. When you said “I could do much better today” in reducing parts count, I think a lot of people would be interested in a parts-reduced design with similar performance.  Any thoughts on NC2030 simplification would be appreciated.  Maybe it’s time for a new kit??
 
Regards,
“Digital Steve”, K1RF
 
11715 2015-10-01 11:35:12 kb1gmx Re: Atlas 210x reboot
I'd create a separate topic for the forum on that project.  Keeps them from all getting munged together.

For SMA at work a torque wrench. 

For the rest of the time a very short 3inch 5/16inch wrench.  over the years I've acquired a 
calibrated finger.  Doesn't hurt there is some latitude.

As to RG174... I hate the stuff.  Melts, fragile, lossy.  I use RG316 PTFE easier to work with
more controlled loss.  For things that need it there is a double shielded version.   Ut141-50
vendable hard line is a handy items too.  I get that from Fleas as there is always some  
cheap to free with SMA connectors.


Allison


11716 2015-10-01 14:26:44 Tayloe, Dan (Noki... Re: Atlas 210x reboot

Actually, I was referring to reducing the number of “unique parts”, not the total parts count. 

 

However I have learned a few things since then.  That is the point in building and experimenting.  If I were to do it again, I would certainly change a few things.

-          Dan, N7VE

 

11717 2015-10-01 19:04:03 kb1gmx Re: Atlas 210x reboot
I understand what you said only that lot of simple parts Rs and Cs are only an 
issue to kitting it (production) and required attention of the person assembling.
Understanding what that part is or does is repeated to the extreme.

If it were so many toroids, they are simple things but winding them scares 
people for reasons I have yet to grok.  If they are each different it speaks to 
skill of the builder. They must be made, terror is often what I hear.

When I said hundreds of ICs, same deal, they weren't all the same,  but the 
markings are on each and it was identify and insert.  That and there were 
large numbers of the same thing repeated.  The complex part was the 
maybe 10,000 plus wire wrap wires all going from some point to another 
point.

I understand that kitted designs seem to appeal more if the total parts 
count is small.  That is also counter sometimes to extending the design
in ways that improve it.  The best example is AGC, I've seen more reasons
to leave it out.  The best was it "complicates the design".  Really?  We build 
a whole transceiver maybe but another 5 or 10 parts (likely less) is a big 
deal.  The addition of AGC means we can have ear saving and likely 
more comfortable experience listening.  

When yet another a minimal design pops out people flock to it and 
often for reasons that are more emotional than design or even best 
practice.  I see that every time I listen down around the QRP part of 
CW or SSB segment where I hear chirpy, drifty, sometimes even 
buzzy CW, and SSB with the VFO or maybe the CIO FM modulated 
or drifting.   All it needed was a bit more of something to insure that 
was not the case.

Excuse the rant. 
Home brew should work well too.  There are plenty of examples 
that do and some cases exceptionally well.


Allison