EMRFD Message Archive 9452
Message Date From Subject 9452 2013-11-29 17:49:59 rcbuckiii Grounded gate amplifier impedance I'm looking at Fig 9.64 on page 9.35 of EMRFD. What is a rough estimate of the input impedance of the output low pass filter (L3,L4,C4,5,6)?
I have entered the values into Elsie for evaluation purposes. With a specified input impedance of 22 ohms, the loss at 14.0 MHz is -11dB. If I specify 50 ohms, the loss drops to -8dB. If I specify 500 ohms, the loss is only -0.7dB.
The input low pass filter (L1,L2,C1,2,3) with 50 ohms in and out shows a -0.5db loss in Elsie.
9454 2013-11-29 18:43:23 Ashhar Farhan Re: Grounded gate amplifier impedance I threw the filter into the GPLA that comes with EMRFD using the 6-8 MHz values.1. I didn't notice any reference to 22 ohms being specified in the text.2. The drain output Z will be in the region of 1000 ohms or more.Findings:1. At 1000 ohms (on the 56 pf side) the loss is quite low.2. The S21 was around 25db. Which is quite good.3. I was also surprised to see that it is actually peaks at 6.7 MHz.- farhan 9455 2013-11-29 22:27:37 Dana Myers Re: Grounded gate amplifier impedance 9456 2013-11-29 22:37:53 Dana Myers Re: Grounded gate amplifier impedance 9457 2013-11-30 07:28:35 kb1gmx Re: Grounded gate amplifier impedance
The amplifier is a combination of a lot of the elements discussed in EMRFD. There is no single page that completely explains its workings. So I will point out those that apply to the fet and also that the section on filters needs to be referenced as they are not only impedance match the serve as peaked low pass filters.
Go back to pages 2.8 and 2.9 (or there abouts) and work out the GM for the fet at the bias level in use and then you can apply 1/gm to get input impedance. Note that for best noise performance the match might not
be optimal for return loss.
Also see pages 6.12/13 and sidebar on page 6.33.
The gain is mostly set by the load impedance and that is selected as desired in the practical range
and that is usually under 12-14db for common gate.
The 22ohm in the drain is for UHF stability and has no impact (insignificant)
9458 2013-11-30 07:48:16 Ashhar Farhan Re: Grounded gate amplifier impedance Allison,
As the output impedance is set by the termination and so is the gain.
If we use this to feed a crystal filter, the reactive termination
could lead to instabilities. This probably explains why grounded gate
amplifiers are so prone to instability?
On 11/30/13, firstname.lastname@example.org <email@example.com> wrote:
> The amplifier is a combination of a lot of the elements discussed in EMRFD.
> There is no single page that completely explains its workings. So I will
> point out those that apply to the fet and also that the section on filters
> needs to be referenced as they are not only impedance match the serve as
> peaked low pass filters.
> Go back to pages 2.8 and 2.9 (or there abouts) and work out the GM for the
> fet at the bias level in use and then you can apply 1/gm to get input
> impedance. Note that for best noise performance the match might not
> be optimal for return loss.
> Also see pages 6.12/13 and sidebar on page 6.33.
> The gain is mostly set by the load impedance and that is selected as
> desired in the practical range
> and that is usually under 12-14db for common gate.
> The 22ohm in the drain is for UHF stability and has no impact
9459 2013-11-30 07:55:18 rcbuckiii Re: Grounded gate amplifier impedance
Thanks everyone for the explanations. I had noticed the peaks when I simulated the 6-8 and 13-15 values. Fig 9.65 tends to indicate the filters would be flat but that could just be due to the large range being displayed.
Allison, your explanation really explains the circuit design and how it functions.
9460 2013-11-30 13:17:53 Dana Myers Re: Grounded gate amplifier impedance 9461 2013-11-30 15:35:19 Russell Shaw Re: Grounded gate amplifier impedance 9462 2013-11-30 15:51:30 kb1gmx Re: Grounded gate amplifier impedance
Grounded gate is generally stable due to its low reverse feedback. However the usual instability
problem associated with it is due to the fact that at UHF the devices still have good gain and many
circuits I've seen are not paying attention to the gate lead (too long and bad things happen) and
that any unintended circuit level feedback will bring us to oscillators. The latter is often due to people
trying to get the maximum gain by using a high load impedance or just poor layout.
I tend to use grounded gate, grounded base, or grounded Grid when I need lower gain with
better than average stability. However in all cases I have to respect that the device used will have
significant gain outside the signal frequency range of interest.
The filter after the fet is less an issue than it would appear. Most crystal filters are reactive at the band edges and even may exhibit significant impedance swings but rarely at 50 ohms. Most filters would be
in the range of 200 to 2000ohms for proper termination. So if you match the amps 50 ohms back to the
filter the reflected impedance is not what appears to the eye without analysis. Even then it may not be
in a direction that can increase the load impedance to unacceptable levels.
Also the specific circuit on 9.35 is between the antenna and the DBM where the match to the DBM is
whats important for best IMD. In a different chapter there is a grounded gate amp design used between
the DBM to keep the DBM loaded to 50 ohms and isolate it from the filters excursions. In this case DBM could be a true DBM or any mixer that is termination sensitive.
The whole prone to instability statement is one I find at odds with experience and general application.
The whole application of grounded gate is low input impedance and high reverse loss the latter being
critical to stability. The key here is input to output isolation and if the physical layout doesn't do that
then n=1 or greater and you have oscillation. That applies to lead lengths, component location, and
ground paths. While a lowly 2n3819 is a low gm device its still capable of gain at uhf and the circuit
must be valid there too.
9463 2013-11-30 18:46:52 rcbuckiii Re: Grounded gate amplifier impedance
I used a J310 and built the circuit as described for the 6-8 MHz unit today. I used my AADE meter to adjust the inductors to the exact values in the table. The amp is built on a PCB I designed. The PCB has component positions for both the 6-8 and 13-15 MHz amps. There are relays on the input and output circuits to switch between the two amps. The 13-15 parts haven't been put on the board yet. I placed the amp ahead of a 7-7.5 MHz direct conversion receiver I am putting together for testing purposes.
The receiver was working fairly well without the amplifier with just a DTC bandpass filter in the front end. The front end mxer is a SBL-1.
I am using a 5-5.5 MHz VFO and mixing the VFO output with a 2 MHz crystal osc for 40 meters and a 9 MHz osc for 20 meters. This mixer is also a SBL-1 followed by a +21 dB feedback amp to get the level up to +7 for the front end mixer LO input.
On 40 meters I had a strong carrier at 7.250000 MHz. I don't know where it was coming from as none of the mixing math adds up to that frequency. There was no modulation on it so I don't think it was an AM station bleeding through. Possibly a signal from a FM station was mixing with something in the receiver. Everything is laying open
9464 2013-11-30 19:09:53 Dana Myers Re: Grounded gate amplifier impedance 9465 2013-12-01 09:14:32 rcbuckiii Re: Grounded gate amplifier impedance
No, I will not use any attenuation. I will just set the output level of my generator to -20 dBm. I will use my KA7EXM power meter for measurements.
I think the AADE meter oscillator frequency varies between about 20 KHz and 100 KHz depending
9466 2013-12-01 11:34:08 kb1gmx Re: Grounded gate amplifier impedance
I suggest staying well under 20dbm more like -25 to -30 to avoid gain compression.
The AADE meter I have runs from as low as a few Khz to over 700khz (measured) though I may go higher with small inductors in the 10nH region.