EEVblog #411 - MiniPro TL866 Universal Programmer Review

[Circlotron]

I've got a set of four newer type Xgpro programmers running on Linux with with Radiomans software and I program M27C801 1Mx8 eproms. Don't ask me why... Anyway, if I have 3 going at once it usually works okay but if I have all 4 going one or more programmers often stop partway. I have a powered USB hub and a 5V 2A power supply. The programming software is version 8.51. I'm a bit scared to update it in case something gets bricked. What could be the problem?

[james38]

Did i understand it right. You connect four TL866II plus device on the same hub?
You connect three at the same time it works but when you connect four then you have trouble when you flash a devicechip (M27C801). I'm right?

In this case it seems that the current is too high when you use four programmers at the same time.
The 2 A current source should be higher. Test it with a 3 or 4 Amp source.

By the way when you flash the firmware of the TL866II you always should do it one by one not 4 at the same time.

A small risk when flashing the firmware of the tl866II always could be happen.
When I do this i cross always my fingers.
Since the new device has no recovery process like the older one.

[Circlotron]

Did i understand it right. You connect four TL866II plus device on the same hub?
You connect three at the same time it works but when you connect four then you have trouble when you flash a devicechip (M27C801). I'm right?

Yes. 100%
More precisely, all 4 can be connected but only 3 programming at the same time.
And I was worrying about updating the programming software that runs on the pc, not the programmer firmware.

[james38]

And I was worrying about updating the programming software that runs on the pc, not the programmer firmware.

OK, forgot to read that you run on linux with radiomans software.
I am not familiar with his software so I don't know how to handle a backup if an update fails. Sri.

But regardless of the update question.
The trouble when you connect 4 programmes is an current problem from the power supply.

When you flash 4 devices the current of each programmer could be raised above 500mA per programmer.
In this case your power supply with only 2 Amp are not enough. Do you have a stronger one?

[Circlotron]

No I don’t have a stronger one but I might put a meter across the power and see if it drops while programming. Or maybe a scope. Actually I should first try it on my windows laptop and see if it works okay on that. That would show if it is a software or a power problem.

[radioman]

Actually I should first try it on my windows laptop and see if it works okay on that. That would show if it is a software or a power problem.

This is a good test. Also do another one: leave only three devices connected to that USB hub and the 4th one to another USB port and do some tests.

[Circlotron]

Okay... connected the programmers to a windows 8.1 laptop. Didn't have the original version 8.51 installation CD so I downloaded the latest version 10.06. It demanded that I reflash the programmers before it would do anything so I did them one at a time and checked they would read an eprom to make sure they were not bricked. Then I set all four programmers going and sure enough two of them quit partway... Standby current = 590mA total,  current while programming = 850mA total, so psu would appear to be large enough.

Now what?

Edit -> The two programmers that failed, I swapped those eproms into the two good programmers.  Aha! These "good" programmers now failed partway... These same eproms will program just fine on my one older TL866CS programmer though. And faster too.  One thing is different. When I select M27C801 eprom the older programmer software shows Vpp voltage = 13.5 whereas the newer one shows 13.00V. Also VDD write = 6.25V and 6.50V respectively. Cant select 6.25V on newer one, only 6.5V. Are VPP and VDD voltages critical?

Edit 2 -> Just updated the programming app on the Linux desktop (wouldn't talk to re-flashed programmers) and plugged the 4 programmers into 4 USB connectors on the front panel of the pc. Sure enough, 2 are okay and two fail partway.  These newer programmers just don't handle marginal eproms very well, as well as being slower than the old ones. Dang!

[flash2b]

Does someone have a mirror for XgproV1006_setup.rar since de XGecu site is >4h to download from. Tnx

[firewalker]

Does someone have a mirror for XgproV1006_setup.rar since de XGecu site is >4h to download from. Tnx

Here: https://we.tl/t-GxjEBlc8xy


Alexander.

[flash2b]

Thx !! Downloaded it in 1 second now.....

[james38]

Cant select 6.25V on newer one, only 6.5V. Are VPP and VDD voltages critical?

The datasheet from ST says for Vpp -2 to 14 V and on Vcc/Vio (except A9) -2 to 7V.
So you are in the specific range.

Do you have the same results when you try to burn with only one programmer?
Maybe the software cannot good handle multiProgramming and stucks sometimes.

These newer programmers just don't handle marginal eproms very well, as well as being slower than the old ones. Dang!

Seems more that it isn't the programmer at all but more the Feature for multiprogramming.
The old one can the newer one not so good.

[ElektroQuark]

It has been a long time I have used my TL866.
Today I have tried it and minipro software (6.85) doesn't load (Linux with Wine 5.4). It worked perfectly with 6.85 earlier.
The only changes had been Wine updates.
Is it a wine 5.4 compatibility issue?
I have the last dll installed.

[firewalker]

Just checked. It fails for me with:

Code: [Select]

0009:err:module:import_dll Loading library setupapi.dll (which is needed by L"C:\\windows\\system32\\user32.dll") failed (error c000012f).
0009:err:module:import_dll Library user32.dll (which is needed by L"C:\\windows\\system32\\usb.dll") not found
0009:err:module:import_dll Loading library SETUPAPI.dll (which is needed by L"C:\\MiniPro\\MiniPro.exe") failed (error c000012f).
0009:err:module:LdrInitializeThunk Importing dlls for L"C:\\MiniPro\\MiniPro.exe" failed, status c0000135

wine-5.3 works fine.

Alexander.

[radioman]

The datasheet from ST says for Vpp -2 to 14 V and on Vcc/Vio (except A9) -2 to 7V.
So you are in the specific range.

These are absolute maximum ratings not typical voltages needed for a proper programming.  You can't program  a chip with VPP=2V for example. The datasheet is clear about  programming voltage: 12.75V ± 0.25V.  The VCC voltage while programming should be (according to the datasheet)  6.25V ± 0.25V. 

I simulated a chip write (i don't have this chip though) with a 150ohm load resistor  between VSS and VPP (~100ma current for the Vpp) and 150ohm for VCC (~50ma). 
The VCC voltage was ~6.9V during write and the VPP was about 13.8V.

Here is a snapshot of the VPP for the old TL866A


And here is the newer TL866II+(XGpro)


Both programmers use the same 50uS Vpp pulse width in the main interface but the total programing time is about twice for the TL866II+(~300sec for TL866II+ vs ~150sec for the TL866A/CS). From the waveform we can observe that for the TL866A a programming pulse is Ton=60us +Toff=20us (80us/cell) and for the TL866II+ Ton=Toff=63us (126us/cell).   

Despite all these differences there should be no problem.
@Circlotron can you hook a single Xgpro (TL866II+) programmer with one of those problematic chips and post what the error is when fail?

[radioman]

Today I have tried it and minipro software (6.85) doesn't load (Linux with Wine 5.4). It worked perfectly with 6.85 earlier.

Good to know, I'll try to investigate what the problem is.


Later edit:
Due to some recent wine version implementations i had to rewrite some code from my usb wrapper.  Should work now with wine => 5.4. Thanks for heads up @ElektroQuark.

[ElektroQuark]

Your new DLL version solves the problem with Wine 5.4 and 5.5.
Thank you very much radioman.

[Shock]

Looking to update my TL866A (original as far as I'm aware) to the latest version, will be running the original software on a Windows box in this instance. Are there any precautions or steps I should take prior to running a later version such as 6.85 or is it a guaranteed bricker? Just need to know this so I can recover back perfectly if I need to etc. I have 6.70 6.71 6.82 6.85 downloaded.

[james38]

As far as i know the latest 6.85 don't brick the device anymore. This was on an earlier Version. I don't know which one.
If the Software means to detect a clone it only give you a popup but nothing more. The device itself runs with 6.85.

[teevee]

It looks like the company has made a new product called XGecu T56 with 56-pins
http://forums.xgecu.com/redirect.php?tid=19&goto=lastpost#lastpost

They are a bit pricey still, but I like the design 

[grizewald]

Looking to update my TL866A (original as far as I'm aware) to the latest version, will be running the original software on a Windows box in this instance. Are there any precautions or steps I should take prior to running a later version such as 6.85 or is it a guaranteed bricker? Just need to know this so I can recover back perfectly if I need to etc. I have 6.70 6.71 6.82 6.85 downloaded.

I hope it works for you.

I updated my TL866CS (flashed to A version) which is probably a clone with the latest 03.2.86 firmware and it screwed the device up. When I tried to read an EPROM with it (using the Linux based minipro software), it looked like it worked, but the output was obviously rubbish. To make sure it was the programmer and not the EPROM, I took a different EPROM out of the multimeter I'm trying to fix and tried to read that again. Same thing, rubbish output.

So I thought, what the hell, I'll downgrade it back to 03.2.69 (which is what it had before), but no dice. Radioman's TL866_Updater reports "Bad serial checksum" and can no longer reboot the device into bootloader mode.

So now I'm hoping that I can use my PICKit2 programmer to flash the .hex that I generated from Radioman's tool and get it working again.

-- edit

And it worked. Now running again on version 03.2.82. I think I'll stick with that!
I also found out the the EPROM I wanted to check is actually blank. Kind of explains why the meter was doing mad things when it started up!

[Shock]

That is good news, not updated mine yet. I might have a play later on today.

[grizewald]

The output I got when I'd flashed the latest firmware looked like this:

38 32 35 34 36 33 35 31 31 33 41 39 38 43 42 42
37 31 44 45 36 30 38 34 31 33 45 37 41 30 44 30
C5 CA D3 B0 71 66 1F 6C 5D 42 AB 68 89 5E 77 A4
F5 BA 83 20 A1 56 CF DC 8D 32 5B D8 B9 4E 27 14
25 AA 33 90 D1 46 7F 4C BD 22 0B 48 E9 3E 8B 0B

repeated every 100 bytes or so for the entire size of the EPROM.

[Circlotron]

Every 128 bytes?

[grizewald]

Every 128 bytes?

Something like that. I forgot to preserve the file and only kept the strange data.

[radioman]

I explained this strange data block here: https://www.eevblog.com/forum/blog/eevblog-411-minipro-tl866-universal-programmer-review/msg2309859/#msg2309859
This is decrypted device serial code block stored (in an encrypted format) at offset 0x1FD00.   

Your serial code was correct so it must be a CS genuine device converted to A.
Here i explained the serial code 'encryption' : https://www.eevblog.com/forum/blog/eevblog-411-minipro-tl866-universal-programmer-review/msg1878068/#msg1878068