Cobra 148 and like Receiver mods

[LoyalServant]

I have a handful of Cobra 148s laying around and someone asked me about the so called "super receiver" mods or "rf gain mod" and so forth.... I guess it depends on who does it and what they call it.

They involve replacing a transistor... moving the old one... changing some diodes ... replacing some caps... etc.
So I set out to do this to a radio... but before I did that I looked at the circuit... the mods being done... and the datasheets to the components and that was enough for me to realize... this is not going to hurt anything but I am thinking the gains are going to be so small as to not make a difference - if there is a gain to be had.

I checked the radio before and after and I saw absolutely no improvement.
I was able to get it tweaked to 12dB SINAD @ .17uV both before and after but no better than this.

Has anyone actually done any of these mods and saw a REAL performance improvement?
I read guys claiming all kinds of outlandish things with these mods.... and I can see absolutely ZERO improvement. If the improvement is there it's simply below any perceivable value.

What's up with it?
Is it just one of those .... placebo mods?
They think it's better so it is?

 

[Dawn]

12db Sinad? Are these units imports with FM?

 

[LoyalServant]

No FM.... You can measure SINAD on AM too.

 

[Dawn]

New one for me LS. Only spec I've used is the standard 30% mod / to cw for a 10db drop for the little AM I do. Use SINAD daily. No longer use the Sinad meter but now use the program in the HP-8920. I don't have the NA AM standard tests card, just the FM. I'm curious what the test inventory is on the AM one is. Need to check with some of the guys I know at an aviation shop to see what they're doing.

 

[LoyalServant]

Well the gear I am using is a Sinadder 3 and an Aeroflex/Marconi 2030.

The Cobra 148 specs I found say the radio is good to 10dB @ .25uV
I can hear the tone way down near the bottom... .05uV... so the receiver is pretty sensitive.
I bottom out at .014uV (-144dBm) on the Marconi.
It's only really intelligible (as the SINAD measurement establishes) at .17uV

I just set out to see if this mythical 5dB I read by swapping a transistor was nonsense... I think I found the answer there I was just wondering if anyone came to the same conclusion.

 

[JayMojave]

Hello LS: There where mods flying around to replace signal diodes in some radios detector circuit. Changing them didn't change any thing for me as I have modified one radio, a 148 with the new waz zo diodes.

There are custom receiver modifications out there from technicians that reduce noise and receiver overload, but probably not published on the net.

Because the 148 has such a hot front end I always add in receiver attenuation with the RF Gain, don't need the noise. As the audio is feed to a "Sonic Cushion" (GE47 light bulb that dampens the audio spikes) and a DSP Speaker.

Mr. Bill Good a near and dear good friend who passed away several years ago, had a modification to many radios that removed the rf front end bipolar transistor and installed a Duel Gate Mos Fet Transistor, and other parts (I believe in the AGC) that worked quit well, reducing front end over load condition.

I modified another 148 many years ago with the Lou's Franklins Crystal IF Filter that also worked very well, narrowing the receiver IF Bandwidth. For a CB Radio it worked pretty good on SSB. The 148 had a good noise cancelling circuit which I consider essential. Others have modified the CB Radio with a ham radio type crystal IF Filter (Kenwood TS50 IF Filter), but way more costly. But it makes a CB Radio a hole lot better.

I picked up a K-40 Speech Processor Mic back in the early 80's, this worked very well making my bare foot Motorola and 148 have a slightly stronger transmit output, as I could tell it made the radios needle move more, what more could you want......

Jay in the Great Mojave Desert..... The Mojave Desert has more to offer than having miles and miles of miles to run out of gas on your dirt bike.

 

[LoyalServant]

I figured there were more than these 2SC2999 kits floating about but this is the most common one.
And yeah, it's BS.

Diodes... I can actually see that doing something in theory but it's not detectable in my tests.

Then you have the guys saying this is an "upgrade" and that the manufacturer is "too cheap" to do it.
In fact... In my brief searches I have found that a lot of these upgrade parts are actually the same cost as the ones in the radio. What it really appears to be is availability at design time - not because something was cheaper or better.

See, most of us engineers design taking some things into consideration such as:
- What is in stock.
- What is available at design time that is NOT obsoleted.
(Just because I can get a million of a part does not mean I will use it. If it's not recommended for new designs then it's not going to be used in a new design...)

I have a feeling that the choice boiled down to one or both of those factors.
I am going to make it work with a 2SC1674 if I can get a million of them vs 100k 2SC2999s and the specifications are virtually identical, as in this case.

I think the fact that these guys making these "mods" don't ever consider that there is more to designing down to a price, but it's also availability and it's not always about saving $500 on a production run.
It often is, but if you can't do the production run to start with what is the point of saving the money?

I am going to keep toying with it some but so far what I can tell the 2SC2999 is just a good replacement for the 2SC1674 - if you happen to have one or the other.

In fact... here is a fella that didn't get anything out of it either - on this very forum:
http://forums.radioreference.com/cb-radio-forum/267457-2sc2999-transistor-not-working.html

What he may have failed to do is realign the front end.
I found that I had to do that when I swapped the transistors - and expected that.
The transistor swap does indeed work - you just don't get any additional benefit.

I am going to dig around in my books...... I am curious as to whether or not I can find a replacement component that is superior.

 

[Snakeradiocustoms]

In Reply

I would like to chime in on this one for sure, Ok Number 1 When you change out tr 14 and tr 19 which are the 2 main transistors in a 148 with the original Sanyo made 2999E version not the D version You get 1 s unit of overall gain in the receiver. The Milliwatt output of the 2999E vs the original 2SC1674 and the HFE is a decent difference on a transistor tester.

I will Do a video and put it on you tube. And I also use a diode for the detector d1 21 and d 22 that I found and like to use after 15 yrs of doing radio`s and testing every diode known to man that makes the receiver sound smoother and cleaner overall as they are a very fast switching diode with very low capacitance and great forward characteristics and It is not what all other people use or sell on E bay and no I will not tell everyone what they are lol.

Then In concert with that I re flow mall the inductors in the radio or the tanks or cans for the lay person out there and a lot of the board to get the most out of the radios and look over and re flow a lot of the radio soldering with a 5x lighted magnifier

I also Upgrade all of the cheap crappy tin coated copper wire to a bigger gauge pure copper wire for all main grounds and hot wires for less resistance to flow overall and I upgrade the am audio amp driver to a better transistor with a better gain and signal to noise ratio. Then in concert with this I use My HP signal generator with a db meter to get optimal peak out of the radio then tune at the end for maximum audible gain of the tone at -117db at 30% modulation with minimum noise when done.

You are absolutely correct when saying if you just change tr 14 which is a 2SC1674 it will raise the noise floor 1 s unit you can visibly see it on the meter.

But for the REAL way to achieve a much better receiver much more needs to be done and it is all in the tuning at the end. I will do a video on both transistors.

Great post loyal servant I agree completely with your statement and the sellers on Flee bay are just praying on peoples ignorance and also I have gotten a lot of feedback from a lot of my knowledgeable customers that prove everything I do to the 148 and 2000 gtl does make a difference when working small signals. Whether or not people want to believe me that is up to the individual Sincerely. Tim, Owner Snake Radio Customs

 

[prcguy]

Swapping out parts to improve receiver sensitivity may actually cause other problems if you don't run some tests to look for them. CB is also in a region between HF and VHF were atmospheric noise will limit the ultimate sensitivity improvement you might make. I'm not a receiver expert but do have some knowledge and experience from building lots of microwave receivers from scratch.

Because of atmospheric noise and band congestion, radios will measure different on a bench with a signal generator compared to how they perform with an antenna. You can measure this my checking the sensitivity of the receiver on a bench with a directional couple attached backwards to the receiver input with the coupler IN terminated and you generate into the coupled port. You will have to increase the signal generator by the amount of coupling factor to get the same sensitivity readings as without the coupler in line.

Then take the termination off the coupler and attach an antenna while still generating into the radio on a quite channel, preferably late at night. If the sensitivity did not change then your atmospheric noise and local interference level are low enough where increasing the sensitivity of the radio might be beneficial.

Then you might want to do some measurements on the receiver front end to see where it might compress. You usually have to disable the AGC and sniff around before the first mixer with a spectrum analyzer while generating into the receiver. You can then measure the approximate compression point of the front end by running up the signal generator until you find the point where a 2dB increase in sig gen level only makes a 1dB change at the input to the first mixer.

Do the same test after any front end transistor swap out to make sure the 1dB compression point is equal or higher than with the stock transistor and you might be able to measure any gain from the new transistor at that point.

One thing to consider is whatever the 1dB compression point turns out to be on the test bench with one carrier from a signal generator will be 16dB lower with 40 carriers of the same level like when an antenna is connected and the band is busy.

Making a bullet proof receiver usually involves having a low noise front end with a very high compression point and the same with mixers throughout the radio. If they skimped on the mixer you can sometimes replace it with an aftermarket high level version from MiniCircuits but it will also involve increasing the LO drive to the new mixer by amplifying it to the required level and so on.

Then there is upgrading or cascading the IF filters and if done right its a really good thing for receiver performance.

So it makes little sense to change out receiver front end transistors or other parts unless you measure how it plays with the rest of the radio under simulated high level multi carrier conditions. If the front end or mixer(s) are run anywhere near compression they will create IMD products that will raise the internal noise floor of the receiver and overall performance goes backwards due to the mod. It may measure ok on the bench and you may need 40 or more channels of high level signals to bench test for problems that will not otherwise show up until an antenna is connected. This is especially true if the radio was of questionable design and on the edge of receiver overload before any mods.

In my research and development days I got to experiance some of the problems caused by a receiver front end with too much gain that made the receiver slightly more sensitive at the cost of reducing its overall dynamic range and making it more susceptible to overload. Same with mixers and all other components in the receiver chain.
prcguy

 

[Snakeradiocustoms]

I fully understand what you are saying in your post but the real world feedback I get Is accurate and I do believe in my heart that what I do does make the radio better overall and so do my multitude of customers hams and no hams. I go way above and beyond anyone else and even to the point of taking chances to improve these radios. So I`m happy that you are an experienced electronics wiz with proper education but I am a simple man and cannot afford high dollar test gear so I am doing the best I can and do believe in my heart I produce a very good end result. and I have based all my opinions on what I get fed back to me by my customers in the real world. So I will continue to do what I do and I do not hold a gun to any ones heads to get work done by me. I have done a huge number of 148 gtl`s and cobra 2000 gtl`s with excellent results and I am at the top of my game. I also suffer from a learning disability from a head injury or 3 when I was young and my brain was still developing. So it takes me longer to understand certain things and I have to touch and feel and see things in the real time to understand them. I have done tons of hours of r and d on these radio`s. I reflow all bad joints re flow all inductors and tank coils etc. etc. and take a lot of time tuning and re tuning a receiver for optimal results with minimum noise. So in saying all this and what I have stated before this will be my final response to this post. People can judge for themselves if they want a radio from me or not based on my real world time at the bench and my attention to detail and customer service and my testimonials on my website. And people can watch me work live at the bench form my website for better or for worse with all life's up`s and down`s in a real world live environment. Yes I get frustrated yes I swear but it is who I am and I do take my work very seriously. 73 to all. Tim

 

[prcguy]

I was not responding to your post or criticizing anything you do, at least not yet. My post was to remind people that increasing CB receiver sensitivity may not be that important depending on your surroundings and making it more sensitive might actually detract from overall receiver performance in some cases.
prcguy

I fully understand what you are saying in your post but the real world feedback I get Is accurate and I do believe in my heart that what I do does make the radio better overall and so do my multitude of customers hams and no hams. I go way above and beyond anyone else and even to the point of taking chances to improve these radios. So I`m happy that you are an experienced electronics wiz with proper education but I am a simple man and cannot afford high dollar test gear so I am doing the best I can and do believe in my heart I produce a very good end result. and I have based all my opinions on what I get fed back to me by my customers in the real world. So I will continue to do what I do and I do not hold a gun to any ones heads to get work done by me. I have done a huge number of 148 gtl`s and cobra 2000 gtl`s with excellent results and I am at the top of my game. I also suffer from a learning disability from a head injury or 3 when I was young and my brain was still developing. So it takes me longer to understand certain things and I have to touch and feel and see things in the real time to understand them. I have done tons of hours of r and d on these radio`s. I reflow all bad joints re flow all inductors and tank coils etc. etc. and take a lot of time tuning and re tuning a receiver for optimal results with minimum noise. So in saying all this and what I have stated before this will be my final response to this post. People can judge for themselves if they want a radio from me or not based on my real world time at the bench and my attention to detail and customer service and my testimonials on my website. And people can watch me work live at the bench form my website for better or for worse with all life's up`s and down`s in a real world live environment. Yes I get frustrated yes I swear but it is who I am and I do take my work very seriously. 73 to all. Tim

 

[Snakeradiocustoms]

Reply

Yes I know you are not attacking me or criticizing me . I never said you were just stating my case. I do appreciate the input and I am full aware of all the ramifications of how and why or why not to do any receiver mods to 148`s and 2000`s. I all ways ask how and where the customers radio will be used and what its primary function is. In some cases I will tell them not to do a specific mod to the radio. I have received many hate e mails and comments on you tube throughout my career but I have also received 1000 times more compliments on how I do radios and my honesty and my values when it comes to business. I have had no less then 30 radio jobs in line for the last 2 yrs. straight thanks to repeat customers and word of mouth and new customers from e bay and you tube So If I do get attacked it just goes in 1 ear and out the other. I have a very thick skin now lol. You all take care man. Tim

 

[LoyalServant]

Nobody is attacking or criticizing anyone... this is more of a scientific study.
Does it work... does it work but have bad side effects....is it BS....

I said this in your other thread.... engineers look at numbers, formulas, specifications....
And I agree, numbers and formulas and so forth do not always equal real world.

I am not the only one to have come up short on the claimed results of some of these mods either.
There are lots of posts from other engineers on other forums that I have found that come up short too.
I have not given up on trying to improve it.... it just does not work as advertised.
On another forum someone claims 20dB improvement or explained it as a "S1 to S4" improvement on a signal. That sounds like a pretty big leap.... and one I am quick to dismiss like perpetual motion claims.

Again... not an attack but your "customer testimonials" don't mean a thing to engineers.
To the focus groups, marketing people, other potential customers yes.... but to me it's all subjective information that I cannot quantify.

I muck around with radios as a hobby. My serious engineering work is not... so this stuff is a hobby to kill time and get some enjoyment out of my daily grind. Not making money... not paid to do anything... it's just me playing with a radio.

What would be nice in all honesty is of you joined the discussion.
Actual observations in the real world is valuable information.

I think we get it... your only doing what people ask you to do.
No harm, no foul...there is an explanation.... fools and their money.
It was here before I was born and it will be here long after I am dust.

 

[Snakeradiocustoms]

Video link

Here is a video link of a 29 lx cobra i did and the before and after results real world. You don`t have to watch if you don`t want to. Thanks.

https://www.youtube.com/watch?v=6kmzGNrIX4s

 

[LoyalServant]

Definitely going to watch this as soon as I get done getting my ears lowered.

 

[LoyalServant]

Because this is worth updating...

First, I did not look at the datasheet closely and see an asterisk on the DC current gain values.
When I ordered some from my obsolete parts supplier they sent me the 2SC2999C.

Obviously not what was used in this mod so I have some 2SC2999E on the way.

That said.. until I get the E variants and try those I withdraw my previous conclusion.

I went back and looked at the DC gain on these C variants and it is higher than that of the 2SC1674 even still. The ones I have seem to be performing as good as the top end of the D variants.

What I did check out just out of curiousity was ACI... and it was somewhat worse than with the 2SC1674.
So if the gain is higher between these components, and in this case between the 2SC9999C and the 2SC1674 there is more gain.... it's not seen and in fact it seems to actually hurt the receiver rather than help it.

I can't see the benefit if someone a mile away transmits 2 channels over and swamps the frontend, thus destroying any chance of having that distant contact.

Will test again with the 2SC2999E but I suspect that ACI will just get even worse.

 

[JayMojave]

Hello LS: Keeping an eye on this tread maybe I'll lean something here.

I need to run down and understand the SINAD measurement process more than what I know now.

What is ACI?

Jay in the Mojave

 

[prcguy]

SINAD is a measurement of signal + noise + distortion over noise and distortion. The typical meters that measure this have a very narrow 1KHz filter that will only pass a 1KHz tone from a signal generator and a very narrow 2KHz filter and circuit that looks at second harmonic of the 1KHz test tone and noise away from the test tone. Internally the meter compares the input containing the test tone and whatever is riding along with it and displays it in a dB ratio.

Most people with SINAD meters do not realized you must calibrate the meter to your signal generator or else it will not be accurate! I forgot the procedure but had to do it on a regular basis working at a large land mobile mfr in the 70 s and 80s. You generate a 1KHz and 2KHz tone directly into the meter and peak or null a pot or two.

I was surprised when I bought my first SINAD meter for home use in the 80s and how far off the cal adjustments were when using with my sig gen. Anyway, its a really good and fast way to peak a receiver for best performance and since the SINAD measurement includes distortion, you can peak IF cans and FM discriminator by generating a signal at the maximum BW the receiver is spec'd for and peak for lowest insertion loss and least distortion of the signal, etc.
prcguy

Hello LS: Keeping an eye on this tread maybe I'll lean something here.

I need to run down and understand the SINAD measurement process more than what I know now.

What is ACI?

Jay in the Mojave

 

[LoyalServant]

ACI is adjacent channel interference.

I have an old sinadder 3 that was off as you describe.
Fortunately these meters are very easy to calibrate.

Dynascan specs from way back when specified the 148 to be ...

Sensitivity
SSB: 0.25 µV for 10dB (S+N)/N at greater than 1/2-watt of audio output.
AM: 0.5 µV for 10 dB (S+N)/N at greater than 1/2-watt of audio output.

Selectivity
AM: 6dB @ 3 KHz, 50 dB @ 9 KHz.
SSB: 6 dB @ 1.1 KHz, 60 dB @ 2.3 KHz

Noting that AM specification.... .5µV @ 10dB
I have not done any sideband testing as yet.
I can get .25µV @ 10dB with the stock radio but before tweaking any of them they were in the
.35µV-.50µV range - so within spec.

What I need to check is the selectivity.
If they specify 6dB @ 3 KHz and 50 dB @ 9 KHz when the receiver is at .5µV @ 10dB then these figures will most likely be a lot worse. So in fact tweaking/modding the receiver for increased gain may actually harm it - the question is how much.

I am still leaning on... transistor is not worth it, and tweaks are probably justified if done right.
It's the selectivity that is going to suffer... a lot.
I can tell you now I was not getting 50dB @ 9 KHz.. more like 20dB.
It was actually pretty bad.

 

[prcguy]

Tuning or changing parts in the front end of the receiver before the mixer should have no effect on selectivity, unless you overload the front end and create IMD in the first transistor. Selectivity is determined by the IF filters.

Another approach to improving a receiver would be replacing the mixer with one that has less conversion loss, and that equates to a lower receiver noise figure and better sensitivity. Only do this if the replacement mixer has an equal or higher IP3 rating and is made for the same LO drive level.

You can also usually cascade another ceramic or crystal IF filter since the IF will have some impedance matching in and out of the filter and an identical filter will usually match well to another like filter. This is where you can improve selectivity by narrowing the IF skirts without affecting the overall IF BW by much. A few companies used to do this like Royce and the radios with dual ceramic IF filters were known to have better adjacent channel selectivity. I successfully added filters to a couple of radios many years ago with great results.
prcguy

ACI is adjacent channel interference.

I have an old sinadder 3 that was off as you describe.
Fortunately these meters are very easy to calibrate.

Dynascan specs from way back when specified the 148 to be ...

Sensitivity
SSB: 0.25 µV for 10dB (S+N)/N at greater than 1/2-watt of audio output.
AM: 0.5 µV for 10 dB (S+N)/N at greater than 1/2-watt of audio output.

Selectivity
AM: 6dB @ 3 KHz, 50 dB @ 9 KHz.
SSB: 6 dB @ 1.1 KHz, 60 dB @ 2.3 KHz

Noting that AM specification.... .5µV @ 10dB
I have not done any sideband testing as yet.
I can get .25µV @ 10dB with the stock radio but before tweaking any of them they were in the
.35µV-.50µV range - so within spec.

What I need to check is the selectivity.
If they specify 6dB @ 3 KHz and 50 dB @ 9 KHz when the receiver is at .5µV @ 10dB then these figures will most likely be a lot worse. So in fact tweaking/modding the receiver for increased gain may actually harm it - the question is how much.

I am still leaning on... transistor is not worth it, and tweaks are probably justified if done right.
It's the selectivity that is going to suffer... a lot.
I can tell you now I was not getting 50dB @ 9 KHz.. more like 20dB.
It was actually pretty bad.