clay

iTrader: (6)

Anyone got any suggestions on trying to figure out my lack of RPM signal from my Megastim 2.1? The following is copied from my help thread on msefi.com:

Well, I've got a MS1V3 I am assembling and a Megastim V2.1. A 12VDC power supply is powering the Stim. I'm not getting a RPM signal only (O2 works fine) during the tests at Step 55. I'm doing the Hall sensor ignition and have the three jumpers installed on the main board.

I've gone through the troubleshooting steps here:
Assembling Your MegaSquirt Step 11
and I have found the following anomolies:
I have 5V coming to the Stim on pins 26 and 28 as I should.
R9 tests at 5k or 10k (depending on which way I test it with my meter), not 39k as it is spec'ed.
I have 5V into R9 and 1V coming out.
R12 (my RPM pot) has the 1V coming in and 0-1V coming out and going to my AD654 (U1) Pin 4.
I also have 5V going into AD654 (U1) pin 8 as I should.

I read on another thread that people get various readings on R9, but they made it sound like they are getting 0-5V coming from their RPM pot. If it is supposed to be 0-1V then I'm thinking my AD654 is bad, but I wanted to see if anyone had any more suggested tests I can try. No scope so I'm limited in what I can look for coming out of it. Thanks in advance for any help!

Braineack

iTrader: (62)

did you do the miata mods to it?

clay

iTrader: (6)

No. The only Miata specific thing that I have done is not installed R7 (in order to later use the Miata temp sender). I tested the CLT gauge in Megatune with R7 temporarily installed just to make sure it worked. I'm stopped at Step 55 in the assembly.

jobambo

Your almost there, a few more wires.... https://www.miataturbo.net/forum/t13676/#post168117

shabaz_yousaf

Hi Clay,

I had a similar problem when a friend asked me to work on their megasquirt -
no RPM with the MegaStim (it was V2.2, but the schematic is quite similar to V2.1). So I googled and found your post.
I replaced the transistor and the AD654 but no change. Then I inspected the design a bit closer.. in my opinion there is a flaw there.
Basically whoever designed it, should have stuck a resistor in series with the Base connection to the transistor Q1. I used a 1K resistor, this worked perfectly. All of a sudden everything worked.

Basically the Megastim design assumed that the transistor has a low gain, which is not a great assumption. With the 1k resistor, then the circuit will work with any PNP transistor.
(To get Vbe >0.7V, just 70uA leakage current is needed at the input. Even though the AD654 should not have so much leakage current, maybe solder flux can cause this, since the output pin is right next to the groung connection - not sure. Anyway, I observed it with two AD654 devices from different vendors).

Thanks,

Shabaz.

Braineack

iTrader: (62)

are you talking about R4?

shabaz_yousaf

Hiya,

I think R4 is fine, just another new resistor is needed. See the attachment,
I used MS Paint to draw a red square where the 1k resistor should go.

To do this, I had to solder in the transistor but without the middle pin of the transistor
connected to the board. Then I connected that last pin through a 1k resistor.
It looks a bit messy on the board, but works.
Without this mod, I noticed that the output voltage was a square wave (as expected), but at a level of +8V (the 9v battery voltage) and about +7.5 V, i.e. was not going down to 0V. This means that the PNP transistor was being left on, through a leakage current in the base (i.e. leakage between pin 1 and ground, either due to solder flux or due to the AD654) being sufficient to turn it on enough. With the mod, the signal is a perfect square wave +8V to 0V.
Another alternative method may be just to reduce the value of R5, but actually the new resistor is a better method, and should always work.

Shabaz.

Braineack

iTrader: (62)

Whoops, i meant to say r5. I'll give it a try and see how it works for me.