EMRFD Message Archive 13418

Message Date From Subject
13418 2016-12-31 15:49:32 Rod H mode mixer frequency limits
Hi All,

r.e. EMRFD H mode mixer with FST3125, described in EMRFD pages 6.48ff:

What is the practical upper frequency limit for the LO?

Would 40 MHz be pushing the FST3125 too far?

Any reports of practical experience would be greatly appreciated.

Best Regards for the New Year,

13419 2016-12-31 19:24:02 bob_ledoux Re: H mode mixer frequency limits
Have you looked here?

H-Mode mixer



13420 2017-01-01 08:59:17 bob_ledoux Re: H mode mixer frequency limits
I've been experimenting with Tayloe detectors which often use the FST3253 quad switch driven by a two bit counter.  Frequencies up to 125Mhz are possible with this switch.

It may depend on whether the switch is driven at two or four times the receive frequency.

This is no joke.  Check out Potato brand semiconductors.  They make some of the fastest chips available. 

I presume you've checked out I7SWX's articles in the EMRFD file section.

This is also good:


13421 2017-01-01 09:31:26 Ashhar Farhan Re: H mode mixer frequency limits
The KISS mixer that I built with J310s worked well into the VHF. My signal source tops out at 160 MHz. The loss had gone up to 9db, but it worked. Triquint made FET switching mixers that work at microwaves.

- f

13422 2017-01-01 10:35:59 Rod comprehensive H mode mixer data


Thank you for pointing out this amazing original research on H mode mixers!


13423 2017-01-01 13:57:51 kb1gmx Re: H mode mixer frequency limits
I've used fets at UHF (432 and 1296) and they work well but at uhf leads count.

The problem with H-mode at higher than 30MHz (maybe 50) is not the actual fet switch its the 
inverter/driver that drives it.  It has a finite propagation delay, Ton, and Toff times so at some 
point you can't turn it on and off fast enough.  They were designed as bus switches so the time 
to go on and off obey logic rules even if the switch can pass 125mhz or faster data with little 
loss.  Raw JFETs like J310 do not have this problem.

The single Jfet used in Rick's converters are very good past 500mhz using good layout rules.
The dual single balanced Jfet switch is also good to UHF save for then leads length and other 
parasitics favor a monolithic design  

13433 2017-01-02 02:35:29 i7swx Re: H mode mixer frequency limits
Hi Rod,

I have been using the H-Mode Mixer with FST3125 and FSA3157 to modify the 2nd mixer on several Icom, Kenwood and Yaesu transceivers with input/output IF up to 70MHz, having LO up to 100MHz. Results very positive, not only with test equipment but with the ears ...

In Files you can see a file I published with different set-up using both type of switches. If using the 74AC86 as squarer you need to adjust the bias for LO going over 30-40MHz. Using LVDS squarers (double for FST3125 and single for FSA3157) you do not need any adjustments, conversion loss will be no more than around -6dB.
Using my H-Mode Mixer version with 2 transformers, the 5 winding transformer should be used for the VHF input or the VHF output. The LVDS squarers terminate the LO on 50 ohm, while the 74AC86 will need a terminating resistor at it's input. The LVDS has a better Ton/Toff balanced switching respect the 74AC86.
I tested the HMM on 144MHz and presented a cl of around -9/-10dB max.

I suggest you to wind your own transformers (my version) as you will not have too much loss respect the 3 purchased transformers version. You can see PA3AKE measurements and reports as already suggested to you by N7SUR.

BTW, the Kenwood TS990 has an FST3125 with 74AC86 H-Mode Mixer (3 transformers version).

73 and a Happy New Year 2017 to all