• 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.

2SC2999 MOD ANALYSIS

I have a Uniden Washington base with a Digi-Scan. Noisy and susectible to Splash big time.

Yes. What I am trying to do is find a transistor with low noise and high gain

superior to the 2sc1730L in a TO-92 package if possible. Might not be.

Now that you mention "8719" chassis. Today will "Shotky diode" this Washington Base.

Going to try this Mod. on the Washington 8719.

Please do not make fun of me if I missed something. At least I try !

I like the radio due to the Digi-Scan. I did a PE Mod. Good talker AM and SSB.

with a stock mic. Will try my other Mics soon when I get the Mic adaptors. Too lazy

to swap the 5 pin male to four pin.

Sort of fixed an issue. With a 2sc2166 driving a 2sc1969 would not deadkey past 2 Watts. Switched to a 1306 driving 2sc2078.
Adjusted bias. Then to spec 4 watts.

HERE is My Receive MOD. Tired of being Splashed by Skip Stations.

You know you have to do something when a Guy in Alberta Canada with a Moon

Raker 14 a few hundred whiskeys is bleeding unto the channel your on.

And your ten channels away ! And two thousand miles away !

Many years ago Many. My first solid state radio was the Venerable Cobra 148.

Purchased from CBC International. 455KHZ filterKit & 7.8MHZ filter Kit.

Made a difference. CBC no longer stocks these kits. My 148 was stolen decades ago.

Going to recreate the 455KHZ filter . With new sharper tighter Muratas 2.4 K

Going to try it. Will post whether it works or not on a later date.

I'm going to use the SSB ladder Filter. 7.8MHZ probably 2.4KHZ from a Cobra 138

that was not restorable. As an IF filter on the 8719 Chassis. My Washington.

CBC would not give me the schematic for the 455KHZ filter. Figured it out.

Its on a Bread Board. The Op Amp. preceding the filter is the basic CBC Op Amp

design of the 7.8 & 10.695 Filters. The 138 did not have F2 to canabalize that part.

F2 on a 858 Chassis base is what helps greatly in preventing bleed over. Alas.Mobile

versions were castrated apparently. F2 is a Monolithic Filter 7.8MHZ 2 Pole.

The filter I will try today is a SSB filter that is quite narrow at 7.8MHZ 2.4KHZ I figure.

Most likely 6 or 8 pole? Alright first coffee . A cigarette. Then try this.

It might need another OP Amp stage after the Filter.

Might have to change the 3904 to a 2sc1730. It might not work at all.

Will Post Back ! Keep forgetting to call Uniden. Doubt it if there are any more.


73s John
 
Well. I realized I confused a PNP version Data Sheet similar to BF763.

BF763 has an NF of 5DB at 800MHZ. Most likely will not due. All that

is left- is to investigate SMD PNP transistors that are of newer design.

I have an opportyunity to get 500 pcs of 2sc1730L cheap. I will just buy it.

Next is to use my Cobra 139xlr 858 as a test bed. Want to play with the "Clipper Circuit". Looking at CBC speech processor currently available.

Could not help noticing the similarity in the 858 circuit. 2 x 914 diodes.

The mic gain in the 139xlr actually controls the Clipping level. My trc457 is strong

and clear with the clipper removed and using a Turner 3B +" Compression Mic".

Fallowing 399's suggested mod with a capacitor value change. But the Defom Power Tweak Mod on the Audio circuit. Which removes the Diodes out of the circuit and changes resistor values and adds one. A no no for a purist "399".

What about. a 12DB Op Amp after the Clipper Filters in the 858 to the audio chip?

Okay I'm sold on the 2sc1730L. My Trc 457 recieve on SSB with Shotky Diodes

sounds quieter and just as sensitive to my Cp2000B maybe better. Am oh well.

Let me finish this 7.8MHZ filter experiment LOL ! The CP200B uses a compressor.

the 858 has it beat already. If it works the above experiment. Will put the Trc457 diodes back in !

2sc1730L it is !
 
IIRC, the 858 chassis already haave the 1730 transistor in them, so there is no reason to change it. However, other radio chassis from Galaxy, Ranger, Cobra,ad Uniden that have later pll chips will benefit.
 
IIRC, the 858 chassis already haave the 1730 transistor in them, so there is no reason to change it. However, other radio chassis from Galaxy, Ranger, Cobra,ad Uniden that have later pll chips will benefit.


Exactly ! That is why I am Buying 500 pcs Smiles. However; I did post somewhere a generic circuit using a pair of PNP low noise hygain transistors in parralel.

Will try it eventually with 2 x 2sc1730L. If I run into issues.

Will learn how to use "Spice" to model the circuit and give me the correct values.

Interestingly I have used 2sc2999. Works in a couple of Radios. Most the original is better. If I remember 2sc1675.

Luckily only have about 20 2sc2999. Nickel each. No biggy.

The Shotkys work though quite well in my Trc457 and several other radios.

Allright back to this 7.8MHz filter !

73s
 
  • Like
Reactions: Shadetree Mechanic
Just wanted to add some thoughts here . . .

It is true that the 2999 has only 2.1db gain/avg over the stock 1674 and has only .8db less noise floor. of the same, there is a bigger picture to it For example, if you add both the .8db smaller noise floor with the 2999's/2.1db gain; the perception is a nearly three ('2.9db') or consistent. Your mileage may vary.

But so many experimenters that do this mod and don't have the means to re-align the receiver, it is questionable that any real difference may be realized in every circumstance. The state to tune/alignment in any given radio isn't equal at any one time, so your results will vary accordingly.

In addition, while using Schottky diodes for the Noise Blanker and the receive circuit has modest but useful gains, the original germanium diodes throughout the rest of the radios other circuits should not be replaced with the Schottky devices. I have gotten a radio once so modified, and had to repair it back to stock. The Schottky diodes many be a much faster switching device, but it also has a higher voltage turn-on point. SO the net result is a radio that has less receive sensitivity overall. IOW, best to just stick to replacing the basic four diodes - and leave it at that. More is NOT better.

Just some of my observations and thought so far.
Interested in any feedback . . . Thanks
 
Last edited:
I didn't go down that road :)

But... people swear by this 2SC2999 business.
It's mostly another gimmick to sell a part and a modification.

It really doesn't work as people expect. They think they are going to get x dB every single time and in every situation they use it in as if this is an absolute thing.

Nobody has demonstrated that it has the magic pixie dust they say it does. In fact, it may hurt more than it helps.
These guys don't realise that you can overdrive the mixer/IF stages doing this and unbalance things.

It's a part swapping thing. Have to swap a part because this new part has a bigger number.
 
Hmm...

This "comparison" will need tweaks between both the 2999 and 5770 - and that is where anything else than the original design - is just not going to work well.

The 2SC2999 for example - sure - works, but when it in itself is the only sub, the Common Base mode goes flat, the Common Emitter takes a knee and requires extra capacitance "tuning" so it is more of a Non-Linear RF amp than a simpler amplifier with good Hfe ratings. It does require "matching". A network. Low noise achievements are when you use it in Common Base mode, that configuration suffers from the heavy capacitance tuning - you will get some gain in noise factor but the tuning swaps are not worth the effort - you don't gain all that much. Remember in a Galaxy you have that L5 and its' coupling/tuning issues.

Did I mention flat? Yes, because once gain is established, there is little room more for amplification - the "knee" region spans just less than 0.2VDC and with AGC actions these days, they are mountains with swinging radios - so once the scent of gain is accomplished - it plateaus - it stinks...so overload is your result.

Now, if it were me, I'd tweak the diodes - Germanium gives a "warm" slope and tone to the signal - but earlier ones or the ones that have aged, get "punctured" in their doping region - lessens their efficiency so they are not as effective as they once were.

Schottky types, well are - just that, low Forward Drop Diodes - but are silicon grain based. Some can introduce what I can hear as a "click" its a type of dynamic noise caused by the quick avalanche of the signal thru the junction - barrier - shot noise. The next stage amp is going to pick that up and pass it along the chain. Great for quick response feedback systems - you just don't want it in the audio path of the signal - else it will work too well for the purpose intended - remember the click, for it does show up.

Don't believe me? Look further into the receivers and the carrier detect sections for AM - note the filter cap on the output. Before it goes to the ANL section. Got into an argument one time about this but the carrier signal turns into a bias - that if put thru a diode like schottky - will need extra filtering - and not just a bigger farad cap across the signal to ground - it's simply reacting to the march of the swing and the lag between the IF section, the AGC resistor - the AGC cap - RF amp and IF stages gain and the whole cycle repeats.

Seriously folks, we are entering into a age where wildly swinging radios require drastic measures beyond typical response feedback principled radios are currently designed for. Mimics SSB modes really...

We are looking at signals with FM and AM components mixed in with variable power for the carrier and our current platforms have to put up with that.

So trying to put the cart before the horse - isn't working too well.

:+> Andy <+:
 
Just wanted to add some thoughts here . . .

It is true that the 2999 has only 2.1db gain/avg over the stock 1674 and has only .8db less noise floor. of the same, there is a bigger picture to it

Robb -

If you read my original post again, you will see that the 2999 has 1.9 db LESS gain than the 1674, so the noise figure improvement will be less that 0.8 db too.

- 399
 
  • Like
Reactions: Robb
Next is to use my Cobra 139xlr 858 as a test bed. Want to play with the "Clipper Circuit". Looking at CBC speech processor currently available.

Could not help noticing the similarity in the 858 circuit. 2 x 914 diodes.


Following 399's suggested mod with a capacitor value change. But the Defom Power Tweak Mod on the Audio circuit. Which removes the Diodes out of the circuit and changes resistor values and adds one. A no no for a purist "399".

What about. a 12DB Op Amp after the Clipper Filters in the 858 to the audio chip

The CBC speech processor is pretty much a copy of the processor used in the 858 chassis. You can give it a try, but I don't think you'll see any improvement over the stock 858 clipper.

I've tried the Defcom mod, and the results were less than spectacular. Removing the diodes and adding/changing the resistors negates the clipping and just changes it into a standard AMC circuit.

An Op Amp after the clipper in place of TR20 is (IMO) unnecessary because TR20 recovers the losses in the clipper circuit and provides enough additional gain to drive the modulation to 100%+. I tried substituting a higher gain transistor for the stock TR20 (2SC945A-Q), but it caused the modulation transformer to be overdriven and produced distorted audio.

To me, the 858 chassis is the best mass-produced type-accepted cb ever made. I bought my TRC457 new in '78 and still run it because, for all-around performance, it can't be beat.

- "Purist" 399 :LOL:
 
So in essence there are no easy answers? Things have become too complex for that? I agree!

Oh well. I just wanted to make a good thing a little better, if I could. I may try the 1730 swap if I can find the location on the board. Old eyes.

Thanks for the help guys.
 
Last edited:
Well, if you bought 20 Sanyo 2N2999'e' transistors for a buck; then that is a really good deal. But - did you test them all for hFe? IIRC, the hFe should be more than 180 - but I know that unit_399 will have the actual specs. It would be one way to see if they are bogus and counterfeits, since transistors bought on Ebay can often be.
 
  • Like
Reactions: rabbiporkchop
Well, if you bought 20 Sanyo 2N2999'e' transistors for a buck; then that is a really good deal. But - did you test them all for hFe? IIRC, the hFe should be more than 180 - but I know that unit_399 will have the actual specs. It would be one way to see if they are bogus and counterfeits, since transistors bought on Ebay can often be.
The cheapest price I could find 10 years ago was about $0.18 each before the supply dried up.
 

dxChat
Help Users
  • No one is chatting at the moment.
  • dxBot:
    Tucker442 has left the room.
  • @ BJ radionut:
    LIVE 10:00 AM EST :cool:
  • @ Charles Edwards:
    I'm looking for factory settings 1 through 59 for a AT 5555 n2 or AT500 M2 I only wrote down half the values feel like a idiot I need help will be appreciated