RM Italy 203P Gate bias voltage?

[doffo]

Hello.

Something I noticed when trying to set the gate bias voltage of the RM Italy amp I was trying my hands on to fix, it seems that setting the voltage of the gate bias to 3.6v, the amp would behave erratically on SSB. The relay sometimes would not click back off, and the relay would sometimes stay on the entire time until you clicked the amp back off. Putting the gate voltage to around 2.2v seemed to of remedied the problem. Amp will now operate as it once did and no problems with the relay clicking back off when you unkey. Am I overlooking something that is causing the problem or is lowering the voltage of the gate to 2.2v not a good idea? I am using the FQP13N10 transistors as thats what I had.
All the help so far is greatly appreciated.

 

[sp5it]

Without temperature compensation it won't work correctly anyway.
If I don't know right bias value for the transistor I set bias for best linearity and IMD on the scope and spectrum analyser.
Mike

 

[doffo]

Without temperature compensation it won't work correctly anyway.
If I don't know right bias value for the transistor I set bias for best linearity and IMD on the scope and spectrum analyser.
Mike

That might be something that I was missing. I was under the assumption that the transistor just needs enough voltage coming in to turn on. From factory I believe it was showing something like 1.2v? Which would be too little for the transistor to turn on? I am probably wording things wrong, but if it is set with a scope, then the amp will have to just wait.

 

[Dr_DX]

I just measured the gate bias voltage on my KL-203 (should be the same as a KL-203p) with the stock RM-3 mosfets. It is showing 1.85v. R3 is a 3.3K to ground that if replaced with a 5K variable should give enough adjustment to get the FQP13N10's working. They have a gate voltage threshold range of 2-4v.

 

[Dr_DX]

In thinking about this, probably a better method to correctly set the bias is to use current rather than voltage. Most TO-220 case N-channel enhanced Mosfets used in radio's and amps for RF require at least 10ma to "turn on".

Put an amp meter that can accurately measure up to 500ma in series with R2 (10K) and measure the current when the amp is keyed. You should see an "increase" of at least 40ma; possibly more. The higher the current, the more "linear" the Mosfet is (good for SSB), but the more heat is generated (put a 80mm fan on it).

If you don't see an increase of 40ma with the FQP13N10's (the circuit is designed for RM-3's), change R2 to 6.8K and add a 10KVR in series with it. You then should be able to adjust the VR to get an increase of 40ma (or more) to make sure the FQP13N10's are turning on fully and are linear.

It has been shown that using the current method of setting the Bias on the Mosfet/s gives better results than using voltage.

Good Luck.

 

[doffo]

In thinking about this, probably a better method to correctly set the bias is to use current rather than voltage. Most TO-220 case N-channel enhanced Mosfets used in radio's and amps for RF require at least 10ma to "turn on".

Put an amp meter that can accurately measure up to 500ma in series with R2 (10K) and measure the current when the amp is keyed. You should see an "increase" of at least 40ma; possibly more. The higher the current, the more "linear" the Mosfet is (good for SSB), but the more heat is generated (put a 80mm fan on it).

If you don't see an increase of 40ma with the FQP13N10's (the circuit is designed for RM-3's), change R2 to 6.8K and add a 10KVR in series with it. You then should be able to adjust the VR to get an increase of 40ma (or more) to make sure the FQP13N10's are turning on fully and are linear.

It has been shown that using the current method of setting the Bias on the Mosfet/s gives better results than using voltage.

Good Luck.

Thank you so much for this information.

I think the current way I was doing it was changing R2 to a 20KVR, and measuring the voltage of the Gate's pin.

I think putting the the multi meter in series and checking the ma would be smartest to set at R2 vs trying to read the voltage.

Even when setting the voltage to upwards of 3v, it would stick the relays and sometimes not even turn back off.

I will report later on when I can have some time to try it again. Thanks!

 

[doffo]

One more question,
Is it best to key the amp by jumping TR1 so that the AM carrier from the radio doesnt mess with the measurement? Unless when it keys up, The ma measurement doesnt get affected by modulation.

 

[Dr_DX]

Yes. Just take a test lead with a "hook" and attach it to the exposed lead of R5 closest to TR1(underneath the output transformer wraps). You can key the amp by just grounding the other end of the test lead. This grounds the low side of the relay coil.

 

[doffo]

Yes. Just take a test lead with a "hook" and attach it to the exposed lead of R5 closest to TR1(underneath the output transformer wraps). You can key the amp by just grounding the other end of the test lead. This grounds the low side of the relay coil.

Thank you so much! Makes the task easier doing it this way. The first 203P is now running good. I will do the same for this other older red 203 in the coming days. I only have 1 more original 203P from around 2011 when I bought it with the stock mosfets in it. Hoping it still runs for more years to come, but now I at least gathered up enough to at least bring it back to operation. It may not produce all the power it once did, but that is ok. Rather a working amp vs a paperweight.

 

[Dr_DX]

RM-3's are what come stock in the 203 and they go for 5~6$ each. If you want to pay less, I suggest the Vishay/IR/Infineon IRF520 or IRF520PBF (PBF=Lead free). They are less than 1$ each and you need to use a matched quad for the amp. If you want a little more out, use the IRFZ24N; the IR/Infineon brand are the best.

I would hang on to the FQP13N10's. They have been discontinued and they are OEM in the President Lincoln II+ - there is 3 in each radio. They will become valuable in the next few years.

 

[doffo]

RM-3's are what come stock in the 203 and they go for 5~6$ each. If you want to pay less, I suggest the Vishay/IR/Infineon IRF520 or IRF520PBF (PBF=Lead free). They are less than 1$ each and you need to use a matched quad for the amp. If you want a little more out, use the IRFZ24N; the IR/Infineon brand are the best.

I would hang on to the FQP13N10's. They have been discontinued and they are OEM in the President Lincoln II+ - there is 3 in each radio. They will become valuable in the next few years.

I mighta lucked out with the FQP13N10s in a set of 4 as thats how I bought the pack of 5. Unsure if they are matched?

Been trying to look up how to go about building something so I can try and make sure they are matched. Would you have a link to build something that would do the job?

 

[nomadradio]

FQP13N10's. They have been discontinued

Mouser doesn't have them any more. It's marked "obsolete" on their web site.

Plenty of sources on fleabay, 12 to 15 bucks for ten.

73

 

[Dr_DX]

https://www.firstwatt.com/pdf/art_matching.pdf

I use one of these and try to match on all 4 parameters with Vt and Cg being the most important.

Amazon.com

You can get them cheaper on Temu or Aliexpress if you are willing to wait.

Try to use the same brand and preferably from the same batch. Even on devices that have the same #, different brands have vastly different parameters.

For example a Vishay IRF520 (OEM in Q6, Q6pro, Q5n, Q5n2, CRT7500, CRT7900, CRT9900, QT-60, QT-80, and others) and a IR IRF520 are very different.

 

[nomadradio]

For example a Vishay IRF520 (OEM in Q6, Q6pro, Q5n, Q5n2, CRT7500, CRT7900, CRT9900, QT-60, QT-80, and others) and a IR IRF520 are very different.

Odds are that all of them make a perfectly-efficient switchmode chopper/oscillator stage. That's the job these parts were designed to do well. The RF properties are altogether an accident. As such, the folks making them have no need to tightly control the RF properties.

Back in the bad old days you would need to adjust the pi-network tank coil in a sweep-tube linear to compensate for the difference in capacitance from one brand of 6KD6 to a different brand of the same tube. The variable "neutralization" capacitor in the hybrid ham HF transceivers like FT101, TS520/820/830, Drake 3 and 4 lines was there for this reason. Changing brands from one set of finals to the next would require adjusting that capacitor. Using replacement tubes that came off the same assembly line would usually make this unnecessary, since the mechanical differences inside the new tube would be very small. Those tubes were meant for use at a much-lower frequency than 1.8 to 30 MHz, so the RF properties inside (like internal capacitance) were not tightly controlled. Wouldn't make a color TV work any better to do so.

Some things just never change.

73