• You can now help support WorldwideDX when you shop on Amazon at no additional cost to you! Simply follow this Shop on Amazon link first and a portion of any purchase is sent to WorldwideDX to help with site costs.
  • Click here to find out how to win free radios from Retevis!

Using a high gain transistor to drive affordable low gain bipolars?

Onelasttime

Sr. Member
Aug 3, 2011
1,229
779
123
So I have a bunch of MRF 458's and HG 2879's that I purchased long ago for $29 each around 16 or so of each. Most of them are matched pairs or matched quads.

I think in my youth I was taught that low gain bipolars where not a problem if drive with a high gain transistors. In my youth high gain American made transistors and Japanese made transistors where common and cheap so this was not a thing I really though about.

So I would like some input on this.

I was thinking about using a high gain American made or Japanese made transistor to drive lower gain American or Chinese made transistors.

I was thinking about using a high gain bipolar to drive say 6 lower gain American or Chinese transistors. It would be a B or AB biased not class C. I would probably make the driver switchable so I can use it with low drive old school radio's and modern high drive. Will likely do Manhattan style build since I only have 1, 2 and 4 transistors PCB's.

I have two MRF455's that depending on the current used to beta test them read high 90's to triple digits. I was have been saving them since the 1990's to service my 1998 President Lincoln should it pop it's final.

I was thinking about using one of them to drive either 6 MRF458 which are basickly low beta MRF454 or 6 HG 2879's. Motorola transistors need very little drive so maybe an all MRF box. Since the HG's are lower gain transistors I am not sure if 1 high gain MRF455 is enough drive.

I am rag chewing type not a com box type.

I know it makes more sense to use the HG's for a build like this since they are nothing special but part of me would like to do a all Motorola box as well for sentimental reasons.

I could skip the high gain MRF455's and use an HG as the driver or use a MRF454 or MRF458 as the driver but they are all lower gain.

In my youth I had fantastic results with 1 and 2 pill triple digit MRF455's purchased for like $9 per transistor in the early 1990's.

1 driving six is about all I want to power current wise any more than that and I might as well build a tube amp!

Thanks for advice, experience and opinions! I have plenty of MRF458's and HG2879's and

This would be the largest amp I have built in terms of transistor count. The MRF parts have all come from RF parts the HG's directly from HG.

I am really only trying to get hot TS667 power levels with out driving the tar out of my gear. I will likely use this for base operation with a combo of vintage CB's and mobil Ham radio's. It will initially be broad band with filtering on the output but if I like it I will build tuned input for each band.

Again Thanks!
 

hot TS667 power levels with out driving the tar out of my gear.
Kinda reminds me of freeze-dried water. The 667V is a DX500 that gets the tar driven out of it. Not sure what makes that desirable, since four of the larger transistor types would be quite happy with the drive from a typical two-final "10-meter" radio.

If the ambition here is just to burn rubber, a driver stage between the radio and four transistors is the way to go. How much it's worth just to see a wattmeter show more is not my department.

And if communication is the objective, leaving out the driver would be my advice. And remember, free advice is worth every penny you pay for it.

73
 
There is nothing about the inherent gain of a transistor that makes it more or less unstable. The circuit they are in does that.

Remember the SRF devices? All those are is double dot black cap Motorola devices. The highest gain devices they made.

Have a high gain device that's unstable? Add more feedback. It's now cleaner, stable and has gain.

There is nothing about high gain driving low or any other such nonsense that makes something stable or not.


The circuit is stable, or it isn't.

Don't believe the hype.

--Toll_Free
 
Kinda reminds me of freeze-dried water. The 667V is a DX500 that gets the tar driven out of it. Not sure what makes that desirable, since four of the larger transistor types would be quite happy with the drive from a typical two-final "10-meter" radio.

If the ambition here is just to burn rubber, a driver stage between the radio and four transistors is the way to go. How much it's worth just to see a wattmeter show more is not my department.

And if communication is the objective, leaving out the driver would be my advice. And remember, free advice is worth every penny you pay for it.

73
Love it "And remember, free advice is worth every penny you pay for it." I have 3 sons and the youngest two are 21 I know how free advice works! LOL You put a smile on my face with that gem of wisdom!

The TS 667 was just a well known point of reference for desired power output with out driving the snot out of the transistors to get to that point. I use mostly older low power radio's or newer low power Chinese SDR. For instance my Chinese made all mode HF rig is a 20W PEP radio. That will not even tickle 4 MRF458's let alone 4 HG's and I am looking at wanting to drive 6. My mobile 29/50/144/430 is no output beast either.

1 low output MRF455 driving 6 MRF458 or HG 2879's was supposed to be not driven as hard as the TS667 which I agree is grossly over driven. Being able to switch the driver in and out is supposed to allow me to use this as a straight 6 device should I have radio with that much headroom and power. Should I want to use a CB radio though the output is not sufficient to drive 6xMRF458's or HG2879's.

I prefer not to daisy chain amps to get desired drive level that is just asking for problems.

If you like freeze dried water my friend you do not know the insanity of the order of the world! A guy in Germany is working on powdered beer! Given the disdain for non-alcoholic beer in Germany and the food purity laws I think this man has issues that can not be solved with a long lever no social engineering!

P.S. If I had taken 1/2 of the advice my Dad gave me growing up I could be ruling the world secretly by now!
 
Last edited:
My power supply could handle more I am just a really conservative guy.

So you think I should go bigger instead of 1x6 I should consider consider doing 8xMRF458 like the TenTec Hercules II or maybe 10 with a driver or two? I would still need a driver though. The TenTec Hercules II needed 65w-85w to get 500w output. I am sure that was due to input pad. I do not own a single radio with that kind of output power hence the need for a driver.

Even my old President Lincoln with MRF455 final could not drive a TenTec Hercules II.

It is not like the idea of driver section in an amp be it transistorized or tube type is unheard of it is or was very common. Even in a radio you typically have pre-driver, driver, and final.

So I did not see a power output curve based on input power for the MRF458 or MRF454 or MRF455. How much drive does each one need to hit saturation? I think a Toshiba 2SC2879 needs 11w per device to reach Toshiba's specified limit for each device. I would want to have the option to have at least 1/2 as much drive as is needed to saturate each device. In this way you have an idiot an almost proof platform.

I have never built an amp with a driver section before all of mine have been straight designs. I did own a TS667 for short time 3 decades ago before I sold it to go off to college. I own some tube amps with drivers but did not build those.

The drive section will be switched though because at some point I will own a radio that can output 100 watts. My twins though will need to graduate college and have good jobs before I see an Icom in the house.

I have already built plenty of 1 and 2 transistor amps in the last couple of years for mobile and base use. I would like to upgrade my output power. I would like to get closer to legal limit device. I do not need to get their all at once but that is a goal at some point.

So my thought was that a 4xMRF458 or 4xHG2879 would not be much of a step up from 2xMRF454 I am currently using. I do not want the headache at the present time of a large beam to make better use of the power I have. So if I am not going to put up a beam larger than my house the only other thing to do is increase output power. That is how I arrived at 1x6 thinking it would be enough increase over my 2xMRF454 to make the build time worth while.


While I have plenty of bipolar transistors I need to order transformers, teflon wire, heat sink, a bunch of jelly bean parts etc.....I do not do this for a living so I am not sitting on a RadioShack that went out of business! I need to plan this out and get resistors and chokes and relays etc....Hence the reason I seek the advice of the big boys on this site! I live in a town/city of 3000 people no electronic store in for hours. Kens Electronics is 3 hours one way if I do not hit traffic or construction! So I have to order what I think I will need from Mouser.

So I appreciate free advice!
 
The DC Beta of a RF power transistor should have only slight influence on the RF personality it exhibits. The reason for a low-beta choice in a radio was thermal stability, best I remember. The effective DC gain (beta) of a bipolar junction transistor rises with temperature. Literally, a hot transistor has more gain than a cold one. More DC-current gain, anyway. The zero-signal collector "idle" current will reach a meltdown point and exhibit destructive "thermal runaway" failure unless you compensate it. The base current coming from your idle-bias source has to be reduced as the temperature rises to keep that idle current level stable. Linking that base current to the transistor temperature is a big deal if you push an amplifier hard. A low-beta transistor should be easier to keep under control, all other things being equal.

73
 
  • Like
Reactions: NZ8N
The complications of a switchable driver stage aside, I'd advise that you go with 8 finals vs. 6. Your input and output combining and matching will be easier to calculate/tune/implement.

If you have all the parts you talk about, don't hobble yourself needlessly.

Spend your time figuring out the bias scheme you'll live or die by..

My .02
 
  • Like
Reactions: NZ8N

dxChat
Help Users
  • No one is chatting at the moment.