# Help sought identifying this RF transistor

Status
Not open for further replies.

#### tomizett

##### Active Member
Hi All,
I'm afraid it's another dull "can you tell me what this is" thread.

I'm trying to identify the pictured transistor in order to repair a piece of equipment where a number of these devices have failed. It's a SOT89 package (also known as SC62 or TO-243) marked 891 <newline> 29P. I have another example from a very similar piece of equipment which is marked 7C1 <newline> 29P, so I'm assuming that the 29P bit identifies the device.

What else can I tell you about it?
It's an RF power N-channel depletion mode JFET, from a piece of equipment specified to operate from 470 to 870MHz. When they are in circuit and working, they are powered from an overall supply of 12V and draw a drain current of about 430mA through their load resistor such that the transistor sees about 8V between D and S. At this bias level, the gate is between -1 and -1.2V (with S being grounded).

What have I found so far?
Nothing useful, searching by any or all of the part number and package.
Assuming that I might not be able to find an exact replacement, I've had a look around for other JFETs that might fit the bill. So far, I've found:

MAGX-000040-00500P (Macom)
MwT-1789SB (Microwave Technology/IXYS)
SHF-0189(Z) (RFMD)
XF1001-SC (Mimix)

I know that it's not much to go on, but what are people's thoughts on the viability of making a substitution? I'm not an RF guy, so I really don't know how application-specific these devices are. The circuit seems to be a single-stage common-source amplifier; this transistor is the only active device in the signal chain, the rest being a passive combiner network.

Thanks for reading and thanks in advance for any advice.

#### Externet

##### Active Member
2SC891 in surface mount package ¿?

#### Mosaic

##### Well-Known Member
Have a look at this...it's a bit long but shows how to sub out an RF transistors with an MMIC. The part looks similar to yours. If u have a curve tracer or a VNA or even a SA you can prob spec. out a decent sub. part.

#### tomizett

##### Active Member
Thanks both.
It's not a 2SC891 I'm afraid - it's definitely a JFET, although it took me a long time to realise it!

Mosaic, thanks - I'll have a look. We don't have much in the way of RF testgear, only an SA, but that's a start. I could probably knock up a crude curve tracer using a scope and a breadboard. I'm not familiar enough with RF design to know what subtle problems could be caused by a mismatched replacement, although it should be easy enough to tell if it's not working at all.

#### tomizett

##### Active Member
Don't think so, but I'll look over the datasheet to be sure. Thanks.

#### Mosaic

##### Well-Known Member
Can u derive the schematic of the area with the transistor showing the input and output and the voltage nodes etc? It might be an amplifier or an oscillator or mixer etc. It influences the part spec. Especially the Ft and voltage limits.

#### Mickster

##### Well-Known Member
Did someone scratch a "2" on the device, or is it a lazered logo?

#### tomizett

##### Active Member
I think the thing that looks like a 2 is just some fluff or fibre from where I wiped it off before photographing - should have cleaned it better.

That's a great video, so thanks for the link - although maybe not encouraging Mr Signal Path couldn't find a direct replacement for his transistors.
I'll try to trace the circuit out tomorrow - it's quite simple, and looks broadly like the amplifier circuits dealt with in the video.

#### Mosaic

##### Well-Known Member
Actually, I have the direct replacement RF transistors for the repair in that video. I purchased a 4432 3Ghz analog & digital sig. gen. loaded with options for about $550 with a similar problem. Turned out to use MMICs and the prob was a lost signal to the burst modulator. Repairs required a 20K resistor and a day of troubleshooting to get a$2.5K instrument going. Ended up with a few of those RF transistors which look at lot like yours. I got them from UTsource on Ebay.

#### tomizett

##### Active Member
Right, I've done a quick trace out of the circuit. I've drawn the amplifier circuit in detail, and included a rough block diagram of how the entire equipment fits together.
Obviously I've not been able to measure out the passives in circuit, so I've only been able to include values that are actually marked - nevertheless, I hope it gives the idea.

I've marked on the DC operating points in blue, and the JFET device that we're trying to identify or replace is indicated in red.

For the complete story, the equipment we're diecussing here is a transmitter combiner, which is designed to drive multiple transmitters onto a single antenna - the manufacturer's blurb is here:
http://en-de.sennheiser.com/ac-3200-ii
Although to be acurate the one I have on the bench is an AC3200 not an AC3200-II

Once again, any information which could help me identify the original part (see post #1) and buy a replacement, or to find an alternative, it greatfully recieved.

PS. Mosaic, I'll check out that eBay seller you mentioned - it's late here now though, so it'll have to wait for now.

#### Attachments

• 12.6 KB Views: 80

#### JimB

##### Super Moderator
I have just had a happy half hour searching for info on 891 29p and got nowhere.

Having had a look at your circuit and the datasheet for the MAGX-000040-00500P (Macom), my best suggestion is that you buy one (or two), solder it in and see what happens.
My best guess is that it will work OK.

JimB

##### Well-Known Member
This one is tricky. I had a look at sites that list SMD markings, I did a parametric search for something similar, I checked the appearance of that M in the impressed marking. Then I had to wonder why you don't ask Sennheiser?

During the parametric search, I got to thinking that maybe this is an older GaAsFet, the kind that used to require the negative gate bias. When I was doing 800 MHz work in the late 80's, those FETs were common power amplifier devices for the 0.2 to 3 Watt range, which I think this is in.

##### Well-Known Member
MAGX-000040-00500P (Macom)
MwT-1789SB (Microwave Technology/IXYS)
SHF-0189(Z) (RFMD)
XF1001-SC (Mimix)
None of these examples were drop-in replacements. The MAGCX-000040... is way off, it has a fairly high impedance output operating point so no good. The other three are a bit closer, but the last two are closest. I think that both would require tuning the gate bias and both would likely require re-working the output impedance match, which would be a pain. It might be easier to drop in an MMIC and remove the input bias and Z matching, and modify the output Z match. But, doing that is effectively redesigning the equipment and what you do will have a big affect on performance (and nullify the regulatory approval, but thats a minor detail). Without the instrumentation to qualify a new design for performance, I don't think that is a good route to go.

#### Mickster

##### Well-Known Member
This is proving very difficult to find, maybe Sennheiser can help as Ron suggests, but I would not be holding my breath on that..
Sometimes, looking at example circuits from a datasheet for another close-by component within the circuit can glean clues, not in this case.

Additional info for this equipment, including PCB photos, from the FCC report:
https://apps.fcc.gov/oetcf/eas/repo..._id=aGAblnZ1QhgtcueCZAPLGg==&fcc_id=DMOAC3200

#### Mosaic

##### Well-Known Member
Ok, the closest I have found that has the power, package and Ft required is this:
http://www.mouser.com/ds/2/678/V02-0049EN_DS_ATF-50189_2013-11-110-908622.pdf
But the max voltage across the transistor is only 7V and that is exceeded by the spec. the OP mentioned.

I think the last one found by the OP is worth a try. XF1001-SC
The SHF 189 can't handle the current. I believe I have a few of these on hand.

#### tomizett

##### Active Member
Thanks for all the interest guys - much appreciated! I'll take some time to digest it.

We've tried Sennheiser, but that was a while ago (this has been sitting on the shelf for a while) so I may try them again. It looks like the AC3200-II is a current product, and I think it uses the same parts, so you'd think it would be a currently available device, but maybe not.

During the parametric search, I got to thinking that maybe this is an older GaAsFet, the kind that used to require the negative gate bias. When I was doing 800 MHz work in the late 80's, those FETs were common power amplifier devices for the 0.2 to 3 Watt range, which I think this is in
That seems right, Ron. My component analyser showed it up as a depletion mode device - the gate runs at about -1.2v (see markings in the diagram for details). Of course, this confused me when I was first looking at this, because the transistors all read short with no power applied.

It's not the end of the world if we can't fix it - I'll just keep it as a source of spares.

#### Hurty

##### New Member
Hi All, We have the same problem trying to identify this device. We have the same unit with one dead channel. Shennheiser (UK) have the device but refuse to supply one or tell us what the device is coming back with the response "only we can fix this" I'll keep trying to find the device as a matter of principle now.

#### tomizett

##### Active Member
Hi Hurty.
I'm sorry to report that ours is still broken - I never did get around to sourcing a replacement - so I can't offer you anything.
My main concern was fitting something that basically worked but wasn't up to spec, mean that we had one unit that didn't perform quite right or (as someone mentioned above) was possibly in violation of regulations. That's not a risk I'm keen to take.

I too have had the passive aggressive attitude attitude from manufacturers who seem to assume everyone else is a moron - even though what they are most likely to do is just swap out a board. It makes me very angry.

Let us know how you get on.

Status
Not open for further replies.