has anyone tried this?

[mstechca]

I used a 74LS164 with a 74HCT125E to make my microcontroller programmer. Obviously it failed.

Seems that the 74LS164 chip doesn't do what it is supposed to do. I think some documentation made a mistake. It seems that the clock on it is just positive activated, not positive edge triggered (like what most manuals say).

But if it is supposed to be a shift register, why does it act this way?

(someone get me a torch, I'm going to burn that chip )

 

[Nigel Goodwin]

I used a 74LS164 with a 74HCT125E to make my microcontroller programmer. Obviously it failed.

Seems that the 74LS164 chip doesn't do what it is supposed to do. I think some documentation made a mistake. It seems that the clock on it is just positive activated, not positive edge triggered (like what most manuals say).

But if it is supposed to be a shift register, why does it act this way?

(someone get me a torch, I'm going to burn that chip )

I've no idea?, but it seems more likely the mistake is your own? - but I was completely bemused by the obscure choice of components to try and build a programmer?. It seemed particularly complicated for such a simple device? - although I don't use Atmel devices, I was under the impression that programmers for them are even simpler than PIC ones?.

 

[kinjalgp]

I was under the impression that programmers for them are even simpler than PIC ones?.

No thats not true. I have designed programmers for both types. Programming 89CXXXX devices is much more difficult than PICs. They have very critical timing requirements and so if your programmer is parallel port based and using Windows based software (I mean running in a multitasking environment) successful programming of the full device will rarely occur.
I made 3-4 parallel port based 89Cxxxx programmers which didn't work quite well. Once in a while I got "Programming failure" because of parallel port timing problems. The one which had DOS version run well but only pure DOS mode, not in the WIN DOS terminal. I can't re-boot to DOS everytime so I stopped using that too.
Finally made a serial port based programmer having on-board 89C51 to program the device. This never had any problems.

And one more not very important point, PICs have consistent pins used for programming. I mean only a common set of pins are used for supplying Vpp, Data, Clock etc. So if you are designing a multiple device programmer your job becomes very easy, while ATMEL devices have different pins for programming on different devices even though if two device have same number of pins (This is not the case with PICs).

 

[audioguru]

Seems that the 74LS164 chip doesn't do what it is supposed to do. I think some documentation made a mistake. It seems that the clock on it is just positive activated, not positive edge triggered (like what most manuals say).

What's the difference?
The clock is DC coupled and can be used from DC to its highest frequency. Whenever the clock input goes to a logic high, the IC is clocked.

Your problem is that the 2k clock pull-down resistor did not provide a 0.8V or less logic low with the up to 0.4mA clock input current through it.
You left no margin for noise immunity. If your tri-state inverters can drive it, use a 1k pull down resistor.