MOSFET's and the 148GTL

[Pioneer621]

I'm not sure on this but I've been informed that the ERF-2030 shows more gain then the IRF-520 in most CB output stages. I personally haven't worked with either transistor other then using the IRF-520 in switching power supplies. I've tried to find a datasheet for the ERF-2030 and had zero luck. There also may be some improvement if C146 and C152 were replaced with small adjustable caps.

Shockwave,
I am not sure the difference between the IRF520 & ERF2030 myself, I mostly use the IRF520, but have used the ERF2030 in the Cobra 29 radios & the performance was about the same between the two from my testing, but again, I am not sure the main differences.

Yes, good idea, adjustable caps would probably work well in fine tuning the input & output.

My MOD sheet is probably not complete for a MOSFET MOD in these style radios, & after volting the final & getting around 28 watts PEAK it did what I needed it to do & again, I don’t really use this radio much so I did not go any further with it, I kept it pretty basic. So I know it can be improved upon.

 

[Shockwave]

This link may help with the MOSFET. http://www.cbtricks.com/miscellaneous/fet_papers/ekl/graphics/erf2030_148gtl_mod.pdf It says it's for the new 148 but the parts look like they are all the same in the old schematic. They use the ERF-2030 and sell a kit to install it. With respect to the modulation limiter, I use to disable it on AM by removing TR-24. I forgot what transistor controlled the ALC on SSB but removing R-131 may bypass both the AMC for AM and ALC for SSB. You'll see the peak power come up when you get it right.

 

[Pioneer621]

thanks for the post pioneer!!!

i guess i didnt see it when i made my last post.

i should have all those parts on hand to do the MOSFET conversion tonight.

it looks like you did almost the same procedure as the EKL notes on cbtricks list, with a couple of differences.
im curious if you tried any of the other steps and then put them back or what.

like, did you jumper R181?

what made you add 68pF to C146?

did you remove C149?

did you put a 33K ohm resistor from the Gate leg to R180?

i noticed you changed the parts values in the "companion part" you made.
i will also be making my own.
did you do this because the ones you chose were more standard values and you had them lying around or was it because they improved the output?

hopefully i will have time tonight to do all this and report back.
if not, i fairly certain i will have it all done by thurs.

cant wait to see what happens!
LC

Loosecannon,

I just like to experiment with this stuff & what ever values I used is probably what worked best at the time & what I had on hand, although I have so many old parts boards from various electronics I could have made the exact parts from the EKL sheet, & I know I probably did to at least try it in there.

Yes, I know I tried all of what was listed to see what difference it made, & if it did not help I put it back the way it was. R181 is I believe a 10 ohm resistor bypassed with a 0.0047uf capacitor, which would let plenty of RF through, I know I tried a jumper there though to see if it made any difference at all, but it did not so I put it back. As for C146; well, from my experience modding radios such as the Cobra 29 & other radios, changing or adding input & output caps can make a big difference in performance, so I tried shunting some different value caps across them & used what seems to help, although the 68pf cap across C146 does not make much difference, brought it up a little, but not much at all. The 470pf across C152 seems to work the best out of all the fixed value caps I had on hand, so I stuck with that. Again, I did all this to my GRANT XL so any of it could be slightly different or yield slightly different results in a 148, I don’t know.

The 33K-Ohm resistor from R180 to the gate of the MOSFET appears to be for sideband biasing, but to me it seems like too low of a value & even when I tried the exact companion part EKL recommends it still sent to much voltage to the gate (if I remember correctly). I try to set the gate voltage to 3 to 3.5vDC at most, which is about perfect for sideband with a MOSFET, which is what there EN-369FN companion part kind of does, but it is a little on the slow side, it needs RF to convert it to DC for the biasing, but it is a little too slow for crystal clear sideband, but is perfect for AM. Adding the separate resistor for biasing keeps the transistor “ON” even when no RF is present on sideband, that way there is no distortion from the transistor not turning on fast enough. With my parts I used for basically the companion part, using a 1-Meg-Ohm resistor from R180 (which is an 8 volt line when in TX) sent right around 2vDC to the GATE in TX mode & seemed to clean up the sideband a little more, but there was a few watt loss in AM mode afterwards, which to clean up the sideband does not really bother me that much.

Yeah, let us know how it goes if you get a chance to add a MOSFET! Have fun & be careful!

 

[Pioneer621]

This link may help with the MOSFET. http://www.cbtricks.com/miscellaneous/fet_papers/ekl/graphics/erf2030_148gtl_mod.pdf It says it's for the new 148 but the parts look like they are all the same in the old schematic. They use the ERF-2030 and sell a kit to install it. With respect to the modulation limiter, I use to disable it on AM by removing TR-24. I forgot what transistor controlled the ALC on SSB but removing R-131 may bypass both the AMC for AM and ALC for SSB. You'll see the peak power come up when you get it right.

I had to get my flashlight out to take a look in there, but whoever worked on this radio before I got it, cut TR-24 (actually Q-24 in the GRANT XL). I have always meant to replace it, but never seem to get around to it, I just try to monitor my audio & make sure the MIC GAIN is set right when I’m on it & never get any complaints, actually one of my better playing radios.

As for the CBTricks ERF-2030 mod; I know back when I was converting mine I checked that out & at least tried most of what they mention too see if any of it helped, & what did not help I put back the way it was & played around with it until I got it doing what I needed & the mod sheet I made was all of what seemd to work best in mine. But anyone working on the MOSFET in the older Cobra 148 may want to try some of it as it may help.

 

[loosecannon]

thanks for the replies shockwave and pioneer.

a little info on removing TR24 in this chassis. DONT DO IT!

removing TR24 upsets the ALC and causes SSB distortion.
pioneer, im wondering if some of the distortion issues you had on SSB when working with your MOSFET mod might have been due to TR24 being gone.

to remove the AMC limiter in AM mode only, the transistor to remove is TR26.
i just remove R131 instead, and if you look at the schematic, you'll see its basically the same thing.

ok, knowing that pioneer's MOSFET mod didnt blow up his radio, and that it is quite close to the EKL sheets, im going to try doing the mod straight off the EKL sheets and see how that works.

after that i will play around with pioneer's changes and see what happens.

when i do the MOSFET mod, i will replace the stock voltage to the final, get the readings, and then add in the transformer mod.

this is getting fun!
LC

 

[Shockwave]

Don't forget to try phasing the transformer the other way and see if it makes more power. One way should show much more positive peaks then the other.

 

[loosecannon]

i am getting the same readings no matter which way i phase the transformer.

doesnt seem to be hitting the 20-22 watt peaks that it did yesterday.
only getting about 15-17 watts PEP at the moment.

i dont have a way of listening to my output here, so i dont know which is correct.

i will say that i first did it like it shows in pioneer's diagram, and now i have it opposite of that.
LC

 

[Pioneer621]

i am getting the same readings no matter which way i phase the transformer.

doesnt seem to be hitting the 20-22 watt peaks that it did yesterday.
only getting about 15-17 watts PEP at the moment.

i dont have a way of listening to my output here, so i dont know which is correct.

i will say that i first did it like it shows in pioneer's diagram, and now i have it opposite of that.
LC

Here is how mine is currently wired & seems to be “in phase” in my GRANT XL like this. During some more testing I noticed it wanting to break up every so often again, so I added yet another 1000uf cap making it 2000uf & it has not done that since & the audio sounds fine, we will see if that takes care of the problem.

 

[loosecannon]

you should really try putting Q24 back in, and cutting R131 instead.

i have never liked the sound of this chassis with that transistor gone.

i assumed that cutting R131 and turning VR7 all the way up were basically the same thing, but i just removed R131 and the modulation jumped up to over 25 watts.

i did the MOSFET mod as it was written in the EKL notes, but i used a 330 ohm resistor in the "companion part" instead of 360; as thats what i had available.

i also did not replace R181 with a jumper, just seemed like that would not make a big difference.

well, it does not work right now. LOL

when i key the mic the meter light dims and i have no output power, not low power; no power.

i am going to try the mod the way pioneer did it and see if that works, but im wondering if i have a bad IRF520.

i bought about 10 of them from mouser about a year ago, and used one of them for this mod.
they are IRF520A's.

going to test the mosfet i guess just to make sure.

back to it!
LC

 

[Shockwave]

Make sure the case of the transistor is insulated from the chassis using your ohm meter with the radio off. It's probably a defective insulator or a bad FET. Shorting the resistor (R181) may give more drive to the FET and is worth at least trying once you get it running again.

The 148 effectively has two modulation limiters. One for AM called the AMC (automatic modulation control) and one for SSB called ALC (automatic level control). Each is controlled by one transistor on the board. There are many ways to disable the circuit from removing the individual transistor to removing a resistor. The end result is the same.

I don't recommend disabling the SSB ALC circuit as this can cause more distortion problems then disabling the AM AMC. The things to remember here are to keep your mic gain at the right setting. If you disable the AM limiter and keep the mic gain under control, you will have a more natural sounding audio quality with higher modulation peaks.

Adjusting the AMC variable potentiometer for the maximum modulation setting is not the same as disabling the circuit. Even at maximum modulation setting of the VR the limiter is still clamping the audio peaks. This may not be a bad thing if you don't have access to a scope to find the setting that provides maximum modulation without flat topping on positive peaks or worse yet, cutting off the carrier on negative peaks. This is the number 1 cause of adjacent channel interference on CB.

 

[Pioneer621]

you should really try putting Q24 back in, and cutting R131 instead.

i have never liked the sound of this chassis with that transistor gone.

i assumed that cutting R131 and turning VR7 all the way up were basically the same thing, but i just removed R131 and the modulation jumped up to over 25 watts.

i did the MOSFET mod as it was written in the EKL notes, but i used a 330 ohm resistor in the "companion part" instead of 360; as thats what i had available.

i also did not replace R181 with a jumper, just seemed like that would not make a big difference.

well, it does not work right now. LOL

when i key the mic the meter light dims and i have no output power, not low power; no power.

i am going to try the mod the way pioneer did it and see if that works, but im wondering if i have a bad IRF520.

i bought about 10 of them from mouser about a year ago, and used one of them for this mod.
they are IRF520A's.

going to test the mosfet i guess just to make sure.

back to it!
LC

Actually I went ahead & replaced Q24 last night. I was sitting there looking at the radio while it was on the bench & figured I go ahead & change it out.

Did you add the 33K resistor when you did the EKL mod?

If so, try the mod without that resistor installed & see if it works.

 

[loosecannon]

ok, MOSFET mod is working now, but power out is lower than the 1969.

the problem with the mosfet was STEP 8 in the EKL notes.
R180 must be different in the new 148GTL models.

i removed that resistor and replaced C149, basically making my MOSFET mod the same as pioneer's mod.

here are the steps i performed.

replace final with IRF520A.
parallel C152 with 470pF cap.
remove L40.
made companion part with 330 ohm resistor, 230K ohm resistor, and 1N4148
diode.

so, the mod does work in the old 148GTL radios, it just needs to be tweaked a bit.

pioneer,
do you think i should try the 470 ohm and 330K resistors in my companion part?
could the difference in values really make that much of a difference?

also, in a previous post you mentioned something about a 1Meg ohm resistor from the gate of the final to ground, but this step was not in your "mod sheet".
do you use this resistor?

i will try paralleling some capacitors on C146 and see what that does, but im hoping that someone can point me in the direction of what part to play with to raise the deadkey up.

LC

 

[Pioneer621]

Playing with the resistor values may gain you a little more power; going from the 230K to the 330K may give you a watt or two, worth a shot playing with.

As for the 1-MEG resistor; all that is for is biasing the MOSFET, & the value may need fine tuned to get you the proper voltage for your radio. The biasing helps keep the SSB clean, & it did, but I did loose a few watts on AM with it in there. I generally try to set the bias voltage of the MOSFET around 3 to 3.5 volts DC at the GATE.

You connect one end of the resistor to the GATE (PIN-1) of the MOSFET, then the other end goes to R180, either side of R180 should be fine, your just adding it to the +8vDC TX line & both sides of R180 should have it. The resistor drops the voltage of the 8 volt line to the proper voltage for biasing the MOSFET. However, the value of the other resistor that connects the gate to ground (the 230K you are currently using) will change the value of the biasing resistor as well, so you may have to experiment with different values.

The capacitor across C146 did not make a huge improvement in mine, but did help a little.

As noted in one of my previous posts, I did not get a ton of power with the MOSFET in my GRANT XL with stock power going to the final. It jumped up quite a bit once “volted” to the 13.8vDC line in the radio; I basically connected it directly to the positive pin on the power jack, which would be using the TP8 purple wire in the 148.

Maybe someone else with experience with this, or may see something else that can be done will jump in with there input, on the MOSFET in these radios.

I don’t remember what mine did with the 1969 final, but I think it was very close to the MOSFET once changed.

 

[Robb]

Side note: All of the IRF520n's that I'd bought w e r e defective!
Go figure...
I know they are v e r y sensitive to static - probably moreso than other devices. They should ship them in anti-static bags - but they didn't for me.

 

[Shockwave]

I've spotted some issues with the biasing on the FET mod. It seems from what I'm observing that the FET requires more forward bias on SSB then the 1969. I'm assuming you have increased VR-8 for the maximum bias setting and it still didn't clean up low power SSB distortion. This is because the thermal tracking diode D-49 is clamping the bias voltage too low.

Do not apply bias to R-181. If you were going to feed bias into the final from another source through a resistor as suggested, you would apply it to L-40. The side that connects to the emitter of TR-37. By connecting bias to R-181 you have bypassed the RF choke (L-40) that prevents the DC bias from loading down the RF drive to the final. That's the main reason it does less power.

I suggest you modify the existing bias circuit to increase the bias voltage while retaining adjustability. First you need to know how to measure the correct bias. That involves breaking the purple wire that feeds the modulated voltage into the final and inserting an amp meter in series with it.

The radio needs to be in SSB transmit with the mic gain all the way off. The goal would be to see around 40 ma forward bias with the mic keyed on SSB and zero output. There are many ways to increase the bias voltage but I was looking for the easiest way while still keeping some thermal tracking, variable adjustment, and solid stability with TR-37 intact.

One might try placing a standard 1N4001 diode in series with D-49 (facing the same way) and place the diode on the heat sink near the final with some thermal compound. Reduce VR-8 fully counter clockwise before power up so you don't risk over biasing the final. Then slowly advance VR-8 while looking for the 40 ma outlined above on SSB TX.

I should point out that this bias modification has not been tested. There is the chance adding one diode may still not be enough forward bias. There is also the chance adding one could over bias the final. If you're worried about burning up the final, put a 1 amp fuse in series with the amp meter during bias testing.