Hg 2879c

Skydiver256 said:

On these transistors should I get a reading between the collector and emitor,also when swapping from the toshiba to the hg 2879c what all has to change exact parts would be most appreciated

Click to expand...

You probably didn't get an answer yet because it's not so simple. The answer to the first part of your question is easy and may set your mind at ease. It is normal when the negative lead is on the collector and positive on the emitter, to see a forward voltage drop of about .600 volts. Anything between .5 and .7 is OK. You should measure an open with those test leads reversed.

The second question has many variables and will not be the same answer for each amplifier. It is not a matter of just changing one specific part for another. What needs to be changed is dependent on what is already there. Unfortunately HG hasn't been honest about the specs of their part. We know it's not a drop in replacement and therefore the datasheet they copied, does not fit their part.

The biggest difference is the output impedance must be higher since this part has been shown time and time again, to produce less power with less reliability than the original. Less maximum power output, means a higher collector output impedance. That could require either one, or both of the following. First, I would try replacing any fixed value output padding cap with a variable, expecting to lower the value from the stock one. The other is to reduce the number of turns on the output transformer by one and see if that produces better output.

You may also want to try and use a variable cap to replace the one across the output transformer primary. Variables could be replaced with fixed values once the optimum match has been found. The same applies to the input transformers although, they will not be off as much and may fall right into place by just adjusting the value of the input padder cap. You'll know when this is right because the input SWR to the amp when keyed, will be a 1:1.

If you need more details, you'll need to post a picture of the inside guts to this amplifier so that we can be more specific as to what parts you'll likely need to address. This will also produce more input from other forum members.

Welcome! I will be watching this thread closely. Here is a link to my thread about my xforce 4 pill where I replaced the dei with hg and then went down the rabbit hole with tuning. Came out of the rabbit hole with an awesome amp that has flat swr on the input and output regardless of dead key or drive level. Many thanks to all the members who participated in my adventure.

https://www.worldwidedx.com/threads/x-force-600hd-smoked.249976/

The problem with the HG as well, at the factory they make changes but don’t tell anyone. On the regular HG 2879 there are at least 4 different tunes now in the last 2 years. The HG 2879C has at least 2 different tunes. Meaning what worked 6 months to a year ago won’t work now. Values are completely different. So it’s best to take Shockwave’s advice and stick trimmers in it and tune. On the old C’s it only took 500pf across input. But the newest ones like 1000pf across input transformer. A lot different than Toshiba value. Still 1000pf across output ( actually the sweet spot is around 1075pf). The output is120 -150pf. Depending on batch, lead length and relay wires. 5pf difference makes a world of difference on HG. They are very finicky. They also require a lot less bias voltage than a Toshiba

Crusher said:

The problem with the HG as well, at the factory they make changes but don’t tell anyone. On the regular HG 2879 there are at least 4 different tunes now in the last 2 years. The HG 2879C has at least 2 different tunes. Meaning what worked 6 months to a year ago won’t work now. Values are completely different. So it’s best to take Shockwave’s advice and stick trimmers in it and tune. On the old C’s it only took 500pf across input. But the newest ones like 1000pf across input transformer. A lot different than Toshiba value. Still 1000pf across output ( actually the sweet spot is around 1075pf). The output is120 -150pf. Depending on batch, lead length and relay wires. 5pf difference makes a world of difference on HG. They are very finicky. They also require a lot less bias voltage than a Toshiba

Click to expand...

So, basically you've identified 6 different variations in HG's efforts to copy a 30 year old Toshiba... I find it very hard to believe that HG retooled 6 times in an effort to copy the 2879. If that really happened, the engineer working on the project should have been fired 6 times too. DEI already tried to pass a VHF transistor off as a 2SC2879. I suggest that if any of these companies put the slightest effort into cloning the real Toshiba, we would have had it done a decade ago.

I still think they are just relabeling NOS VHF transistors as 2879's. There were just as many of those made, as 2879's and not a single CBer ever got to blow them up, depleting remaining stock, until now. Remember, way back in the 1980's, Motorola made VHF transistors that worked very good on HF like, their MRF492. In any event, we very clearly have learned the new 2879's do not match the specs of any Toshiba 2SC2879. That means if there was any honest attempt to clone the 2879, HG failed 6 times over. It would be interesting to measure their plagiarized output collector capacitance specification, to see just how far off it may be from the "copied" 700pf?

Having to drop the value of the output cap from 680 or higher, down to 120 to 150pf, is a massive change on the output. Having to go from 330 to 1000pf on the input, confirms these parts are extremely different on both input and output. Placing the counterfeit 2SC2879 label on these parts is false advertising. Something that is becoming more prevalent with equipment marketed for this band. I'd also like to thank Crusher for sharing his first hand experience regarding the capacitor values.

We have to remember that Toshiba lost every legal market they ever sold their 2SC2879 in when Mitsubishi came out with their last bipolar 2SC3240 and their new RDXXXHHF1 line of 12 volt RF MOSFET's. If it was no longer economical for Toshiba to continue with their existing part, it's unlikely any other company could recover the start up costs and long term profit, required to begin making the real part from scratch again. The only profit maker DEI and HG have found in this area, is getting a stamp made so they could mark other (lesser) transistors, with the 2SC2879 number.

I don't think hg failed 6 times. I think they implemented cost savings on their end six times.

Shadetree Mechanic said:

I don't think hg failed 6 times. I think they implemented cost savings on their end six times.

Click to expand...

If that "cost savings" idea had anything to do with selling off existing stock by misusing a 2SC2879 label, I'd agree. Since most VHF transistors are more expensive than HF ones at the same power level, few would have ever been considered for use on HF, until now. Other than the MRF492, I've never seen a data sheet specked out for both bands. This means that someone has to take the time making an educated guess to extrapolate HF specs from VHF and then test the part in a real HF circuit for performance.

If the parts are really improving in their ability to produce watts on HF, someone is finding a better existing part. The issue that worries me is when they hit the point where the newest ones are worse than the last ones sold. That would indicate they have already sold off the "other" parts that work best in this application and are now having to use parts with true specs, that are further away from the 2879. Using a 100 watt VHF low band transistor on "10 meters" could easily produce a reliable 120 watts on HF.

If the company was lucky, they may have produced two 100 watt VHF parts with different gains. How long before they have to drop down to the 80 watt NOS parts? If it weren't for the fact the package style changes, they'd probably be restamping UHF transistors, with the 2879 label soon. That would require several more circuit changes to get working (poorly).

Funny thing is they haven’t changed batch code in awhile. So theirs no way of knowing what you get. So every time you get a new batch, you have to build a pair on a tester and check the tune. Just to see if Tune is same or if more changes have been made.

Before watching the video I knew that was the test circuit. The only thing wrong is you never cut the PC board in half, when working with a transistor that has its grounded tabs, coming out the sides of the case. No other part in the circuit is more dependent on good RF connections, with a double sided PC board under it.

Crusher said:

Funny thing is they haven’t changed batch code in awhile. So theirs no way of knowing what you get. So every time you get a new batch, you have to build a pair on a tester and check the tune. Just to see if Tune is same or if more changes have been made.

Click to expand...

Umm... To change the lot number, means they have to buy a new 2SC2879 stamp and that would probably double the cost of their total investment into this project.

This last batch I bought. Was fixing a straight 6 pill/ transistor amplifier. Set up as 2 device sections. Had to match every capacitor ( all inputs within 1 of, all capacitors across transformers within 2pf, capacitors from collector to ground all matched within 1pf and same for outputs) so no imbalance, all new matched transformers ( yes I had to do 2 or 3 wraps on each and check inductance to make sure right material and all very close, within 0.1uH of each other) All new transistors matched. Still have an imbalance. Even though they should all be the same. I tested each 2 device section individually. Bypassing relay and 3 port splitter/ combiners. Each one did different output, vswr on input all different. The regular HG are garbage. I will never use again. The C’s are better. But not a lot.