Legit' Motorola 2N4427 found on eBay

[MiXiN]

Alright necks,

I ordered some of these the other day, they arrived today, & can confirm they're the real deal & not those shiny clones everywhere that are marked "00xx".

These are dated 8435, so manufactured in August 1984.

Motorola ceased production of the 4427 in the 90s, so anything dated 00xx (like clones are) is a fake.

The item # is 233731576578

He sells a long list of new old stock Transistors, so you might find something you want there that's the real deal.

Just thought I'd share & give the heads up, as I've read on here that they're hard to get now.

[teckniqs]

£2.50 isn't bad at all.

[Bton-FM]

The fake ones have a batch number of ‘KO’ followed by 3 numbers. The Chinese 2N4427, ‘5109, ‘3866 are all the same transistor!

If anyone has a good idea for another more modern way of doing the a 1 watt stage that’s cheaper, I’d be interested. I tried the BFG135, but it’s quite hard to squeeze a comfortable Watt out of one.

I’ve got a few others in mind, but the smaller LDMOS devices run at low voltages which is a PITA!

[MiXiN]

£2.50 isn't bad at all.

Even cheaper if bought in multiples bud.

I think I paid £10 for 5 of them.

[jvok]

The fake ones have a batch number of ‘KO’ followed by 3 numbers. The Chinese 2N4427, ‘5109, ‘3866 are all the same transistor!

If anyone has a good idea for another more modern way of doing the a 1 watt stage that’s cheaper, I’d be interested. I tried the BFG135, but it’s quite hard to squeeze a comfortable Watt out of one.

I’ve got a few others in mind, but the smaller LDMOS devices run at low voltages which is a PITA!

I've been playing with the RD01MUS2 (not RD01MUS2B which is a totally different part). Like you say it's meant for 7.2V but it runs fine at 12V. Rated VDSS is 40V. I get about 1.2W out of it at 12V direct into a 50R load without any matching. Probably closer to 1W once I add LPF.

[Bton-FM]

The fake ones have a batch number of ‘KO’ followed by 3 numbers. The Chinese 2N4427, ‘5109, ‘3866 are all the same transistor!

If anyone has a good idea for another more modern way of doing the a 1 watt stage that’s cheaper, I’d be interested. I tried the BFG135, but it’s quite hard to squeeze a comfortable Watt out of one.

I’ve got a few others in mind, but the smaller LDMOS devices run at low voltages which is a PITA!

I've been playing with the RD01MUS2 (not RD01MUS2B which is a totally different part). Like you say it's meant for 7.2V but it runs fine at 12V. Rated VDSS is 40V. I get about 1.2W out of it at 12V direct into a 50R load without any matching. Probably closer to 1W once I add LPF.

How hot is it getting? Also, how much power are you driving it with?

[jvok]

How hot is it getting? Also, how much power are you driving it with?

Not hot enough that I'd worry about it. I soldered the tab down to a solid ground plane which seems to keep it cool (helpfully the tab is the source pin so you can do that). I set the bias for 100mA and just use a diode in contact with the top of the package for temp compensation. With that simple setup bias only drifts by about 5mA between cold and fully warmed up.

Drive power is hard to say yet because I don't have input matching fully worked out. But if you believe the graphs in the datasheet it should be around 20mW. Maybe slightly less because the datasheet numbers are for 520MHz. Gain should be a bit higher on band II.

[radium98]

2n4427 somewhere i saw that 2n4427 work @175 mhz max and another say it go up to 400mhz and here they talk at abt 500

[Albert H]

The RD01MUS2 is a cracking little beast! I found that I could run it at (pretty much) a Watt out with just a Colpitts oscillator driving it. I used a 7808 to run the two stages, and have a lowpass Soshin filter after it. It makes for a really tiny, two-stage exciter!

The second version I tried uses a pseudo-Kallitron two-transistor oscillator using a couple of surface-mounted FETs (like 2n4416). This gives the advantage of isolating the modulator more from the 1 Watt stage and so less prone to hum. The PCB (hand etched, using marker pen resist) including a PLL using an SM PIC and a SM 145170 ended up being just 5 X 2 cm! I used SM moulded chokes for the coils, and spent nearly 45 minutes trying to find the last 22p NPO SM cap I had when I fired it across the office from the tweezers! When I get this cracked, I'll work out a small housing for it, and get it made in the Far East for next to nothing.

[reverend]

Those Soshin filters claim an insertion loss of 3 dB - don't they get mighty hot with 1 Watt running through them?

[Albert H]

I haven't seen 3dB loss through them. Measured here, I lose 1.5dB @98.4MHz, with the unfiltered input at 1121mW (30.5dBm), giving a net output power of ~29dBm at 805mW, and no measurable harmonic content at all. There's also no heat to mention.

I'm experimenting with a rather cunning tracking harmonic filter in an effort to get a truly monotonic 1 Watt VCO out of a tiny PCB in a copper-clad box. Early results are promising, and when I (finally) get home, I'm going to do some serious work on this. I can envisage a cheap commercial module with a built-in PLL, which will be one-time programmable to a specific frequency. (I think that frequency agility is really somewhat over-rated!).

Now that the bigger FETs are getting cheaper, and the broadband circuits are now well-known, good rigs will be getting cheap to build again! The most expensive parts will be the power supplies and the heatsinks!

[Bton-FM]

@Albert H
Are you using varicaps in your filter elements for the tracking?

Also, is the Oscillator affected much by changes of the loading of the RD01MUS2 since there is no buffer? Since the RD01 is MOSFET the OLPD switching could be done by pulling the bias down I ground. Is this how you’ve implemented OLPD?

Is there a benefit of running the RD01 at a lower voltage?

Did you use those coilcraft moulded coils? I saw them a while ago but they seem very expensive for what they are.

[thewisepranker]

So now the cloners are going to mark those BC108s with 8436 instead of 00xx, happy days!

[Albert H]

@Albert H
Are you using varicaps in your filter elements for the tracking?

No. It's a high speed power op-amp configured as a voltage tunable filter.

Also, is the Oscillator affected much by changes of the loading of the RD01MUS2 since there is no buffer? Since the RD01 is MOSFET the OLPD switching could be done by pulling the bias down I ground. Is this how you’ve implemented OLPD?

Not at all. The output from the VCO is from the low impedance point (and purest output) from the emitter of the oscillator transistor (or the source in the FET version). OLPD just grounds the drive at the gate of the FET - it's crude, but there's no significant leakage,

Is there a benefit of running the RD01 at a lower voltage?

Yes. It runs much colder. You don't get as much gain or power out, but it doesn't matter much. The difference is just a few mW.

Did you use those Coilcraft moulded coils? I saw them a while ago but they seem very expensive for what they are.

I used something similar from a manufacturer in Shenzhen. They're not too shabby, and a lot cheaper.