• 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!

DEI Transistors

As we all know, the day of the Toshiba 2sc2879 has come to it's twilight. They are about gone from RF parts inventory, the price of them is skyrocketing, and there is a void left by this. At Xforce, we have moved to the DEI devices to fill that void. There has been some bad press about the DEI units, but we have found them to be a very good device. Yes, if your going to run your system at 21 volts and try to make your whatever pill do however many crazy watts on the line, they may not be for you. but that accounts for about 0.01% of the people out there.
So, keeping with the Xforce Unfair Advantage Philosophy for our customers I built a new gain testing system to check transistors AT FREQUENCY. And I made a video of the test results. I'm sorry for the quality of the video, I used my silly iPhone.
Xforce Gain Testing DEI & Toshiba 2sc2879's - YouTube

XForce.
 
  • Like
Reactions: 1 person

I was actually thinking DEI deserves a lot more bad press. Like as much as they can get for taking the best bipolar transistor number and deceptively placing it on an entirely different and substandard part. Unless quality control is non existent (probably) I wouldn't expect gain to be the trouble. It's the breakdown voltage of the internal junctions and the current handling ability that tells me this ain't no 2SC2879. It's as good as Bofang placing the Icom name on a $39 HT.

Having said that, I admit there is no other part being manufactured today that closely resembles the real 2SC2879. Just don't treat us like an idiots and place a Porsche 928 turbo emblem on a Volkswagen bug while assuring us it's the same thing. If DEI was close to being honest, they would have ripped off the MRF-492 number for use on this part rather than 2SC2879.

I would like to see less opinion and more facts. Just rip the cap off of a bad DEI (should be easy to find) and compare the size of the silicon junction inside the transistor to the Toshiba. Since both parts are rated for 12 volt operation, the only way to handle similar power is to support the same current. That is determined entirely on how wide the internal junction is and how many fine gold wires are installed to carry the current.
 
  • Like
Reactions: 2 people
It's all about the headroom and durability the transistors offer. You're seeing the best possible results with DEI in this video and it's important to understand just about no one is ever going to get the same performance in the vehicle. The transistors are operating just under 16 volts. The point where the likelihood of a failure increases dramatically.

Who runs 16 volts in the car? No stock system is going to produce it and no aftermarket batteries are available to charge at this voltage anyhow. Even a 6 and an 8 volt battery are going to need more volts to charge. The Toshiba could run at up to 22 volts and produce about 250 watts per device before it even reached its maximum collector dissipation (we won't mention IMD). Try that on a DEI and watch it DIE.
 
I know for a fact that DEI recommends that you run their transistor at 90% of the maximum rated power of the Toshiba part. As long as you stay within the stated design parameters every thing will be fine.
As far as Xforce goes, I work on amps day in and day out and I have worked on one for every 20 of another brand I won't mention. If I were going to run an amplifier for just 10 meter it would be Xforce. But I don't run just 10 meter, 160m through 70cm.
 
I know for a fact that DEI recommends that you run their transistor at 90% of the maximum rated power of the Toshiba part.
I don't know what DEI recommends because I've never seen the spec sheet, but here is what RFParts says ..........

RFParts said:
NOTE: Please follow these guidelines when using this part:

MAX VOLTAGE 14v.
Typical power is 90% of original Toshiba part.
Cannot mix this part with original Toshiba. If replacing parts in circuit, you must replace all the parts with same manufacturer.
 
  • Like
Reactions: 2 people
Xforce Gain Testing DEI & Toshiba 2sc2879's - YouTube

After looking at this Xforce advertisement I have a lot of questions about the "unfair advantage" and this gain testing. The DC gain figures are the transistors HFE or beta specification. Signal gain at RF is specified in decibels or db. That's an entirely different unit of measurement that does not require any top secrete circuit or equipment to determine.

All that's required to measure the signal gain of the transistor is a properly matched amplifier and decent watt meter. You simply drive the amplifier within the linear region and measure the drive power versus the output power. Simple online calculators will convert the wattage increase into db. With 4 watts drive you should see about 65 watts output and 13 db gain on one Toshiba.

Now lets get rid of more smoke and mirrors. Bipolar transistors such as these experience a reduction in gain as frequency increases. How does Xforce show about 8 times as much gain on what I assume is 27 megahertz than they do at DC? 6 at DC and 47 at RF? It's impossible to increase frequency this much and not have gain roll off accordingly.

I have no problem using the DEI part today only because there is no other alternative for existing circuits. The problem is even suggesting this part is the same much less stealing the same part number. Look what honest US distributors for these transistors like RF parts have to say about DEI.

RF Parts has published application notes for the DEI that strongly suggest drive levels be reduced by 10% as compared to the Toshiba specifications. The Toshiba can handle 200% of its rated drive and still survive while making more output. I also suspect the Xforce "RF gain testing" drove the DEI transistors into gain compression to get them to appear much closer.
 
Last edited:
  • Like
Reactions: 1 person
I was actually thinking DEI deserves a lot more bad press. Like as much as they can get for taking the best bipolar transistor number and deceptively placing it on an entirely different and substandard part. Unless quality control is non existent (probably) I wouldn't expect gain to be the trouble. It's the breakdown voltage of the internal junctions and the current handling ability that tells me this ain't no 2SC2879. It's as good as Bofang placing the Icom name on a $39 HT.

Having said that, I admit there is no other part being manufactured today that closely resembles the real 2SC2879. Just don't treat us like an idiots and place a Porsche 928 turbo emblem on a Volkswagen bug while assuring us it's the same thing. If DEI was close to being honest, they would have ripped off the MRF-492 number for use on this part rather than 2SC2879.

I would like to see less opinion and more facts. Just rip the cap off of a bad DEI (should be easy to find) and compare the size of the silicon junction inside the transistor to the Toshiba. Since both parts are rated for 12 volt operation, the only way to handle similar power is to support the same current. That is determined entirely on how wide the internal junction is and how many fine gold wires are installed to carry the current.



The Chinese have never played by the rules.
 
  • Like
Reactions: 2 people
Oh, where to start? I have been wanting to write a response to this thread, (the OP, or original poster) for quite some time now, but I have debated on weather or not it would start an argument, or just be disregarded by “CB MYTH” type logic, if you can call it that. (the logic part)

I mean, according to the video, we have an “unfair advantage” and “pure science” working against us here. Ok, enough with the useless sarcasm, if I continued, this post would have the same value as that video, close to zero.

First of all, let's start with the basics..there are several differences in the transistors that can clearly be seen using a simple test fixture (these things the secret tester must have missed):

Let's start from the base of the transistor forward, shall we?

Section A:
1: Input impedance is different for a set amount of RF drive.
2: Input impedance is different for a set amount of drive at a given frequency.
3: Input impedance curve across a change in frequency is different vs the Toshiba.
4: Input impedance curve for varying drive levels changes at a different rate than the Toshiba.
5: Bias current is different for different temperatures than the Toshiba for different conduction angles and collector (idle) current.

Section B:
1: Output impedance is different and follows a different curve for power output.
2: Output impedance follows a different curve for a different amount of VCC. While this is normal for a BJT to exhibit this behavior, it is different than the Toshiba.
3: Current requirement is different for a different amount if Pout at same VCC.
4: Gain is different...we will elaborate on that in detail in a moment.
5: Gain compression happens MUCH more quickly with the DEI transistors.
6: Idle current is different for the same VCC, and bias supplied at the same temperature of the device.
7: IMD performance gets much worse, much quicker with the DEI.

Now, let's talk about gain for a moment. (see section B, reason #4) In the video, the person behind the camera is stating that the reason the DEI transistors have a lower gain, and have gotten bad press (as if this was the only reason, I am not sure why they focused only on this aspect, one that we are about to blow to smithereens) is because that the gain was compared at DC. Ok, while the transistors will have a gain curve from the lowest frequency they are “capable” of and the highest, the statements that the DEI transistors have lower gain at DC and higher levels of gain at 27Mhz is just pure nonsense. Gain goes DOWN not up as you go up in frequency with these devices. In fact, in most HF amplifiers that cover the lower bands, you have to take that into consideration when you design the thing. It is true that certain devices will have a “sweet spot” but this isn't the case here. Then to go on and say that they are close to the Toshiba at 27Mhz is just flat out wrong based on all sorts of facts.

The fact is, the DEI transistors can be loosely at best called a substitute device, not a drop in replacement, and definitely NOT an exact replacement.

While it is true that in a zero bias amplifier, (CLASS C), that is mono-band, (11 meter CB in this case) is such a simple amplifier design, and where things like linearity, IMD, and other desirable traits of an amplifier are not important to the operator, it would be true that one could "fiddle" around with the circuit a little bit, and get the devices to play nice because a lot of the above mentioned could go unnoticed.

So we arrive at the conclusion:
A: The amplifier builders KNOW this, and don't care.
B: They don't know, and don't have the skill-sets to know the difference (most likely the case based on what I have seen)

Is it any coincidence that we see things like a "motor maul" which actually a very simple design, things like unregulated power supplies being used in RF amplifiers, and other horrible slapped together nonsense only in the CB world and no where else? It's because they can get away with it.

Why, instead of trying to convince people that these cheap chinese transistors are just as good,.... why not come out with something impressive? If it's about the money, think about it... an amplifier manufacturer can EASILY build an amplifier that produces a ton more output for cheaper than they would be buying up these DEI transistors for. They would also have a higher profit margin as well. There are many devices out there that are mosfet based (NO-NOT THE SAME RE-PURPOSED SWITCHING POWER SUPPLY ONES, I MEAN REAL RF DEVICES) that can withstand excessive (I'm talking 30:1) SWR without blowing, require VERY simple bias circuits, can be driven hard without blowing, and can be ran on 24-30 volts which would be easy to set up in a mobile using the same techniques that their "motor maul" boxes use to come up with the voltage. You can even build this device yourself, it is very easy. (which is why I believe they figured the concept out)

What I don't believe they have figured out, is how to build an amplifier that resembles anything other than a cut and copy design of the same old push pull bipolar transistor amplifiers using broadband transformers and combiners that has been around since the dinosaurs and put it slapped together in a big metal box on a big metal copper clad board. Ever see their bias design? It's decades old, and a poor attempt at the Motorola EB63 bias method which wasn't even supposed to be used in the real world, but only for testing. There are tons of better ways to do it, but I have never seen it in a CB amplifier from these guys. I am not sure of in the video, if you look closely, that note that it written that says “test your sh*t” is codeword for something else or not, but it must be codeword for something else. No reputable manufacturer using secret science testing methods of the trade that are some type of new testing technique on a BJT (bipolar junction transistor) that has existed for over half a century.

Do I seem to be targeting them? Yeah, admittedly a little, but they asked for it when they started being self-proclaimed experts for the purpose of separating unknowing people from their money.

I understand that they are an amp builder, and they have to pay their bills...I get it. I also get the fact that they have a problem to deal with, seeing that the Toshiba transistors are not made anymore and when you can get them, it does not make sense to pay the dollar amount for something that will cut into your profit margins when you can simply blind users from the truth and call a chinese substitute “just as good” or “equivalent” to what you were using before. BUT, in my opinion, it would have been better not to say anything, other than to post this ridiculousness and be called out on it as well as making claims that their (cringe) “research” is based on science, and there are no “smoke and mirrors” here. This is just begging for someone, like myself that would normally not care to bother, to clearly do everything possible, using no hidden super-duper-secret test equipment to prove, using real world data and test results, using recognized name equipment to show just how bogus the claims are.

Oh, and on a side note, I would be more than happy to post a couple videos if need be, with all equipment out in the open, no secret, unfair advantage, the wizard of oz, man behind the curtain BS.
 
Oh, and on a side note, I would be more than happy to post a couple videos if need be, with all equipment out in the open, no secret, unfair advantage, the wizard of oz, man behind the curtain BS.

Feel free to go ahead, any testing and info would be welcome.


73
Jeff
 
Agreed. I am in a transition at the moment, but will make a point to also show some tests as far as idle current, IMD at different frequencies both RF and AF, different VCC, current draw, I'll even sacrifice a DEI to show what happens at 14 volts into a not so great reactive load at different phase angles.
 
  • Like
Reactions: 3 people
Radioguy, very refreshing seeing more people use what they know to inform others. Nice change from using it to manipulate the unknowing into forking over cash for what's really no advantage at all. Well put and informative.
 
  • Like
Reactions: 2 people
I would be more than happy to. The only thing is, I'll sacrifice the DEI as stated earlier, after it goes "pop" I will also pop literally the case off and take a few snap shots. The only thing is, I really don't want to sacrifice a Toshiba, so if someone has a bad Toshiba 2sc2879 around maybe they could take a photo to compare it to. I would if I had a stash, but the bad ones got thrown away a while ago, and my new ones, well... I really don't want to sacrifice one.
 

dxChat
Help Users
  • No one is chatting at the moment.
  • @ Joseph23:
    Happy Thanksgiving to all and a very merry Christmas and a happy Newyear also. !!!!!!!!!!!!!
  • @ Moleculo:
    Congrats to GlocknSpiel for winning the Black Friday Radioddity QT60 Pro giveaway contest!
  • @ Deon:
    Good day, I have an Icom IC910H radio. I want to program a memory with split frequency tone on TX only and NO tone on RX,. Is this possible to do from the face of the radio? I don't have the interface to connect to a computer.
  • @ crappykraco:
    @Deon very unlikely to get a response here. You are better off starting a thread on the topic.
  • dxBot:
    Backwoods40 has left the room.