“Fake“ Z80s?

[Martin Giese]

I ordered some Z80 CPUs from China, and they arrived today.  They are marked Z84C0008PSC and both of them work in my computer at 7.3728MHz.

Their behaviour differs from Z84C0008PEG I got from Mouser (just in time before they were sold out) in two respects: 1. the power consumption of the whole system is about 100mA higher, and 2. it pull signals only up to about 1.3V below VCC.

Am I right in suspecting that they sold me some rather old NMOS CPUs, rather than CMOS as advertised?

Martin

 

[Tom Pipes]

Hi Martin,

Same here.  I ended up with some lower end Z80 CPUs that had been remarked after running them on my z80 tester.  Worth it to buy from the reputable vendors, fortunately I wasn’t out too much buying them cheap in my case.

Regards,

Tom Pipes

Sent from my iPhone

On Dec 10, 2024, at 1:19 PM, Martin Giese <mabar...@gmail.com> wrote:

I ordered some Z80 CPUs from China, and they arrived today.  They are marked Z84C0008PSC and both of them work in my computer at 7.3728MHz.

Their behaviour differs from Z84C0008PEG I got from Mouser (just in time before they were sold out) in two respects: 1. the power consumption of the whole system is about 100mA higher, and 2. it pull signals only up to about 1.3V below VCC.

Am I right in suspecting that they sold me some rather old NMOS CPUs, rather than CMOS as advertised?

Martin

 

--
You received this message because you are subscribed to the Google Groups "RC2014-Z80" group.
To unsubscribe from this group and stop receiving emails from it, send an email to rc2014-z80+...@googlegroups.com.
To view this discussion, visit https://groups.google.com/d/msgid/rc2014-z80/9b7746f8-7f2c-4b24-b4f4-b215eaec69bfn%40googlegroups.com.

[Doug Jackson]

It is *always* worth lodging a claim with the seller when they remark chips, especially if you have evidence.

The only way of stamping this practice out is to make it uneconomical.

Kindest regards,

Doug Jackson

em: do...@doughq.com

ph: 0414 986878

Follow my amateur radio adventures at vk1zdj.net

To view this discussion, visit https://groups.google.com/d/msgid/rc2014-z80/DC08A0F1-431E-47CA-98CC-B5E1BFD81C07%40gmail.com.

[Tim Radde]

I bought a bunch from Jameco before they ran out.  Not that I need that many but wanted to have some spares.

[Martin Giese]

I also checked the undocumented OUT (C), 0 instruction (ED 71), and indeed, the parts from Ali put 0 on the port while the CMOS part sends $FF.

 

I’ll ask for a refund.  Ali usually doesn’t argue at least, but they’ll want to have a video of me swapping CPUs in my computer.  Not exactly hassle free.

Ironically, the parts they sent me are more "retro" than what I ordered.  Should I be happy? 🤔  At least now I know that it was worthwhile to use HCT logic throughout.

Just checked, Mouser has a couple hundred 20MHz parts left. Pretty expensive though!

[Sergey Kiselev]

It is very common for these Chinese refurbished Z80 CPUs to be not what they are claiming to be, this includes marking NMOS CPUs as CMOS, as well as labeling them with higher frequency rating than they were initially rated for.

(The problem is obviously not limited to the CPUs, and to be honest, the Z80 CPUs is probably one of the best case scenarios, as most of the time you get something that works like a Z80 CPU... With more rare or expensive ICs, many times people get something in a similar package... sometimes even not in a similar package... that is a completely different part, and doesn't work at all)

Back in the spring, when Zilog announced discontinuation of Z80 family, I wrote a CP/M utility to test Z80 CPUs. It is far from being complete, but most of the time produces the correct results:

https://github.com/skiselev/z80-tests

Sample output:

A>Z80TYPE /D
Z80 Processor Type Detection (C) 2024 Sergey Kiselev
Raw results:       CMOS: 2C U880: 00 XF/YF: FF
XF/YF flags test:  C000C000
Detected CPU type: Zilog Z84C00

[Bill Shen]

Here is a recent thread about relabeled Z80.  In summary, vast majority are relabeled, you can distinguish NMOS from CMOS by power consumption; the difference is about 100mA.  CMOS, regardless of labeling, are generally good to 25Mhz, some up to 30+mhz.  NMOS have greater variability; most are good to 10mhz, but few are only good at 2Mhz.  ‘Good’ means passing CP/M’s zexall instruction tests.  Z80 is indeed one of the better yielding used CPU; by comparison, the majority of used 6502 are nonfunctional.

https://groups.google.com/g/rc2014-z80/c/xOLlGqN5Tdw/

[Martin Giese]

Thanks for the pointers!

Will try the z80-tests.

Well, at least they are working Z80 CPUs.  I ordered a batch of 3.3V LDOs from Ali once and got what turned out to be a collection of 78xx of different voltages.  Questionable NPN transistors labelled as JFETS.  A power MOSFET that was a 7805.

On the other hand, I got a bag of C3355 transistors (data sheet rates them at 6GHz) and they work perfectly in a 433MHz oscillator. Good luck trying to find a through hole UHF transistor via reputable sources.  Or a batch of MC10216 ECL parts, most of which worked.  It’s a strange world.

So I try to only order things from Ali when I think I have the means to test them within a few days of their arrival.

[Martin Giese]

Made a little video illustrating the increased supply current and 4V high level to prove that my parts are NMOS. AE agreed to refund within 12h of me opening the case.

[Doug Jackson]

If I get anything that has been resoldered or relabelled, a dispute is instantly flaged and the seller is called out.

--

You received this message because you are subscribed to the Google Groups "RC2014-Z80" group.
To unsubscribe from this group and stop receiving emails from it, send an email to rc2014-z80+...@googlegroups.com.

To view this discussion, visit https://groups.google.com/d/msgid/rc2014-z80/fcb68963-296e-42ec-9164-bbfcced5a76fn%40googlegroups.com.

[Lee Adamson]

Coming over here from that "why can't I set double speed mode on the z180 thread", so as to try to stay on topic...

Upon closer inspection I am pretty sure these have been blacktopped.  I can't see well enough to tell with the naked eye, but the line where the black top ends along the beveled edge is more apparent under the camera.

How annoying!  Makes me wonder what else I have that is fake!

[Bill Shen]

Look at the backside of the chips.  If chips have the same markings and date code on the topside, then the backside should look exactly the same.  Sadly, just about every parts you got from China are relabeled.  Z80/180/280 relabeled parts should work mostly, but consider the labels for information only.  You need to test them as soon as you’ve received them.

Bill

[Tadeusz Pycio]

The forgers have once again amazed me. They have taken the time to create a new designation which is a combination of two different Zilog ICs, the Z8S180 and the SCC chip. They are probably selling this counterfeit chip as Z180 and elsewhere as SCC, as I can find no other explanation for their making such a ‘mistake’.

[Lee Adamson]

Well, the black-topped ones I tried seemed to run ok at 18 MHz at least.  There are vague differences in the cases though, so I guess I am going to have to make some kind of test jig and label them as to what type they actually are and what speed they will run at.  :/  I've had them for several years, originally bought to assemble a Z180 accelerator for a TRS-80 Model 2.  Never realized they weren't what the claimed to be.  :(

The yellow ones don't show any evidence of black-topping that I can tell and all look the same on the bottom, but they came from the same ebay seller, a guy in California.  The first one I tried was dead though.  Second one booted right up and has been running a-ok at 36MHz for the past hours.

I'll have to be more careful in the future.  For the most part I've had good luck with buying this kind of stuff from ebay, but I filter it to "north america only" even though it costs a little more that way.  Oof.

[7alken]

;