mosfets

[bushwacker]

So here is one that I have been pondering over in my mind... These mosfets (erf2030) by ekl that seem to becoming fairly popular latly have interested me since they came out a while back.
This is my thoughts on the subject.... In most circumstances fets are used in an exact oposet way than how they are used in a cb. In a standard cb the final runs off of low voltage and high current....... pretty much the exact oposet way that fets are used in most circuits--- alot of the time fets are used with a high voltage low current situation.
Although they do work with low volts high amps, there seems to be a problem with the need for alot of drive on the fet to get any power out. Like in a standard Cobra 29... usually the outcome of a final switch to FETS is a lower carrier high modulation outcome--- and the stuggle seems to be getting carrier power out of these--
I'm sure everyone has alot of different mods for these for bringing the carrier up some--
My idea is to run the fet like many others are.... higher voltage lower current
Basically stepping up the voltage after the driver, rematching the input impediance and taking the current draw/volt draw off the audio IC and adding it from another source ect ect.
The main idea is to try running the erf 2030 and final rf amp circuit of a cobra 29 off higher volts lower current like how alot of fets are used. I am very curious as to what the carrier/ modulation output would be.
This is just an idea I've been playing around with in my head for a bit ...I'm curious if anyone else has any thoughts or idea about using these fets this way

 

[DTB Radio]

The thought is ok, but woukd entail some major redesign, as well as major new design. In the FET mods that I've done (cobra 29's primarily), I see carrier levels of around 5 to 6 watts, with my Wavenode system regularly recording 40 to 45 watt peaks. I've had customers come back after having the EFT mod done and tell me they walked on guys running amps. The only things I do differently from the cbtricks procedure is use higher cap values, change the driver base input resistor to 4.7 or 3.9 ohms, and add a 33pf cap to the spot where the 1k resistor is removed (i forget the number right now, its near the antenna socket by the SWR bridge circuit). Oh, I usually replace that 1k with about 15k so there's still bleed-off for any static electricity that might be building up.

 

[thetnhillbilly]

My little search came up with this. It is not mine, I haven't tried it, and I can't remember the source, but I would love to hear from someone who has tried it:

"I do have some new info for you on the Cobra 29 / Uniden mod. The existing Cobra 29 mod is very popular, but the only complaint I have ever heard about this mod is that the dead-key is low with the ERF-2030 (approx. 3-4 watts with the current mod).

We have come up with a couple of additional steps to the current Cobra 29 mod that increases the dead-key to approx. 8 watts without affecting the peak power.

To get the higher carrier, do the current mod and then do the following 3 additional steps:

1) Add a 4.7K ohm resistor (1/4 watt) from the positive side of C116 to the gate of the ERF-2030 (this is the left-most leg when looking at the face of the ERF-2030).

2) Remove the capacitor at C62, and replace it with a 390pF capacitor. Also, make sure that there are no additional capacitors tacked on the solder side of the PCB (sometimes Cobra adds additional capacitance at this part to get everything to tune-up correctly). If there is an additional capacitor on the solder side at C62, then remove it.

3) Remove diode at D8, and replace it with a jumper. Or, just add a jumper wire across D8.

We have had good success with these additional steps and it gives everyone the carrier level that they are looking for."

 

[bushwacker]

5 or 6 watts clean is pretty good... I've got around that number but I had to volt the predriver which I really don''t care to do.. unfortunatly I didn't have long to play with it. I did get in some more fets and I'm going to check out what I can get.

To run the fet off of high volts low current would certainly be a total rework of the rf amp section ( voltage supply ect ect), but its one of those thing I would like to try just to see what these things can actually yeild.
My thinking is that more carrier power and perhaps swing can be achieved from running the fet off high volts low current.

DTB.... thanks for the reply. Any thoughts or ideas you have for my project let me know.
Thanks again

The thought is ok, but woukd entail some major redesign, as well as major new design. In the FET mods that I've done (cobra 29's primarily), I see carrier levels of around 5 to 6 watts, with my Wavenode system regularly recording 40 to 45 watt peaks.

 

[thetnhillbilly]

Like I said, it isn't my mod, but it was what I found in a search for higher carrier. Thanks for the feedback. Step three is bad then. How about the first two steps? Are they a good way to get the wanted carrier?

 

[bushwacker]

Thats one thing that I don't care to do in any am radio...... jumper the anti spiking diode. its not worth it for the 1 watt gain (or so). I have done it in scap radios that I was playing around with.... but if I ever did that in a customers radio, the head tech at the shop would more than likly consider shooting me LOL.
:shock:

 

[bushwacker]

and add a 33pf cap to the spot where the 1k resistor is removed (i forget the number right now, its near the antenna socket by the SWR bridge circuit). .

Thats a new one I havent heard... I guess you mean R55 ( I think thats the one) or R54.... anyhow I know what you talking about.... replace the resistor with a cap-- cool I try to remember that one
Thanks for the reply

 

[DTB Radio]

and add a 33pf cap to the spot where the 1k resistor is removed (i forget the number right now, its near the antenna socket by the SWR bridge circuit). .

Thats a new one I havent heard... I guess you mean R55 ( I think thats the one) or R54.... anyhow I know what you talking about.... replace the resistor with a cap-- cool I try to remember that one
Thanks for the reply

I personally don't remove that resistor, just change it to a higher value (about 15k-22k). It is R55, by the way.

 

[W1R3T4P]

I have had a Uniden for 3 years with the spike diode jumpered.
Even ran it at 16v. What is the problem?

 

[DTB Radio]

The only problem I've ever encountered is with Cobra 29's. On some 29's, with that diode jumped, strong receive signals at high volume settings can begin to garble severely, and sound like harsh squealing.

 

[thetnhillbilly]

DTB:
You had said a while back that
"I've been using the ERF2030 final, using the basic mods listed on the cbtricks.com website, with 2 exceptions. I increase the cap going from the driver to the final to about 150pf, and the cap that goes to ground after the final to 100pf. Other than that, I follow whats on cb tricks."

Do you still do this, and then also do what you listed in this thread to get the higher carrier, or do you not add the 150pF and 100pF anymore? You seem to have a great handle on the erf-2030 mod and I'm curious as to what you found to work best.

 

[DTB Radio]

I still use the 150 cap. Most of the time I see 5 watts key, give or take half a watt. If the key seems really low, I just add a bit of bias to the gate of the 2030 by running a resistor from the 8-volt TX source to the 2030's gate.

 

[Switch Kit]

from reading what you have all said on these mosfet mods , I truly have a better idea on just what your talking about here after trying one on a Cobra 25 LTD . Now I have a lot of respect for CBTricks and there site , but I surely think there's a lot more to these mosfet mods then the simple little layouts they put on there site for the hobbiest. They mention nothing about the radios (25s/29s) basically coming up with 1 to 2 watt DKs after the mods ? nor do they reccomend on how to fix the problem as you guys have talked about in this thread. Now I can understand the folks that call these mods JUNK ! it's not there faults that there's technically more to it then is written in there simple little mod instructions on CBTricks . One things for sure with my limited knowledge on these mosfets , I surely like the 1969 mods compared to these, up to this point. I think a little bit more instructions on these mosfet mods on CBTricks would surely be more honest to the hobbiest who would like to get it right the first time and not wonder why things were not working out after they followed the instuctions in the first place and then starting to second guess themselves for basically no reason other then there was surly a lot more to the mods then was put in the instuctions in the first place. Like I said .....I can surely understand when somebody says these mosfets suck when there's a lot more to the mods then are written in CBTricks instructions.

 

[ken white]

MOSFET's are nonlinear devices when used in a circuit designed for large gains. Since a MOSFET is a voltage controlled device, vice a current controlled device like the BJT, it will not load the bias circuit as much if it is just substituted for a BJT. This means the transconductance (gain) will probably increase with this new operating point change. And finally, MOSFET's have parasitic capacitances between the gate, source and drain which has undesirable effects with high frequency.

So, the bias circuit and coupling circuit will really need to be redesigned to get the correct operating, gain, and high frequency response desired - a MOSFET is not a dirrect drop in replacement for a BJT.

 

[bob85]

ken its been a long time since i was at colledge (25 years) but i remember they taught us that mosfets where high impedance voltage gain devices and when used in an amplifier typically produced 10x the distortion of a bjt stage with higher open loop gain and different gain/bandwidth which resulted in the need for more feedback to reduce gain and distortion plus a revised bias supply,
they also said fets have transconductance droop similar to a valve/tube amplifier which made them less prone to thermal runaway than a typical bjt amplifier, something was said about the need to avoid self oscillation but i dont remember what it was, maybe it was something to do with the capacitance which i guess would also effect bandwidth,

have i remembered what i was taught or do i need to get in my parents loft and find my old schoolbooks??
i have not looked at mosfets since i was 15 years old lol,
i am sure i could learn something if i read my old schoolwork.