Who is running MS with a Bosch ISV?

Can't find a direct link but if you go here, then click on "Download datasheet" you'll see it. It appears to have an integral reverse diode (RHS of last table).

Couldn't comment on op-amp suitability I'm afraid.

 

I had the same problem with the ISV when I did my MS install. I messed around with loads of different frequencies, tried restricting the amount of air into the ISV, tried 3 different ISV's one from a G60 and two from a 16V and then came to the conclusion it was just not gonna work! It seems like there's just too narrow a band of resolution for these VDO/Bosch ISV's to work well. One of them also blew my TIP 122!.

In the end I decided to use a standard Ford ISV (off a Ka) and machine an adapter plate for it to bolt onto the end of the inlet manifold. Not only does it work BEAUTIFULLY but looks much neater than having all the pipes connected to the VW one.

Looking at the internals of the the Ford ISV, it works in a similar way to the Bosch/VDO ones but has some kind of bellows that dampens the the oscillation. As a result it's possible to see the bleed valve move in discreet steps when driven from MS. I cant remember the frequency I drive it at off the top of my head, but are more then happy to check if anyone's interested.



Cheers,
Andy

 

interesting but I can't see how I would hook it up to my G60.

I'll perserve with Trevors mod.
RS or maplins don't seem to sell the SSP5N90's though..

 

Andydavy, that's another very interesting solution to the problem. I really like how you have installed it - that's very neat indeed. You must have done a fair bit of experimentation to stumble across an ISV that has the right mechanical characteristics to work nicely with the 'raw' PWM straight from the MS.

I'll stick with the solution I have now for the standard Bosch ISV as the work is now done, and I like to keep things OEM. I keep admiring how neat your solution looks though!

Robert, you'll be able to use virtually any N-channel power MOSFET to be honest. I just looked at the Maplin site and they have:

N32AH IRF740 80 in Stock 1.21
N10AH IRF540A 3 in Stock 1.21
N00AH 2N7002 SMD 134 in Stock 0.13
UF89W 2N7000 153 in Stock 0.44
N93AG IRFZ24N 47 in Stock 0.69

Choose whichever of those is a TO220 case style N-channel MOSFET that's able to handle a few amps.

 

Ok.
This weekend I built a very simple pwm cct.
It has a FET on the output to supply enough current on the output stage to drive the bosch ISV.
Thats it.

At about 144 hz at 87 percent duty cycle the car starts every time with no use of the throttle and maintains a nice smooth idle at about 1000 rpm.

If you disco the isv from the car anbd look at it I can see that the valve is part way open.
With MS the valve just open fully no matter what settings i set.
B4 I use Trev's invention I want to further investigate what is so different in my MS configuration as i have proved this vale will work by just pwm'ing it.

 

oh dear
my previous post dosn't read too well.


oh well.
Ok I have got the car working just using megasquirt and the TIP120/121/122 mod as per the MS manual.
My idle pwm graph is quite flat,but the car starts without throttle everytime now.
It is running at 160hz and the isv is nice and quiet.
So it can be done.
The key for me was seeing that it could be done using my separate cct.
I then re-connected MS and started the car.
It did its normal thing where it rev'd the car very hard to 3k and rising . However instead of panicing and turning off the engine I just used the tuning window to lower the pulsewidths to a value that brought the rpm to a nice normal value.

Hope this helps someone else down the line.

 

What PWM resolution does the MS2 provide at 160Hz? I am using MS1 with MSNSE firmware as you know, and the PWM duty resolution is inversely proportional to frequency. At 100Hz you can get 1% resolution, but at 200Hz it is as poor as 2%, and so on. It is impossible to control the Bosch ISV at 150Hz with such poor resolution when you're driving it directly with switched voltage PWM, and that's where the current ramp circuit solves the problem. But I reckon that your MS2 must be giving you much finer PWM resolution to play with, thus enabling you to drive the ISV simply using voltage switching PWM.

Cheers

Trev

 

tbh I didn't measure it Trev.
I guess it dosn't matter to you now anyway but If your curious put your microcontrolled isv driver on your car and get the car running to that so that you can see what I mean.

 

Have you got it sorted Trev?. I`ve re-witten the idle code complete and added an RPM based ignition trim/advance when in idle mode. Took 6 months on and off to code but it`s rock solid.

My firmware`s a country mile away from the MSNS Hi-res code I used as a base. You can have a look at it but your car probably won`t run with it (I use the Digi1 TPS switches instead of a TPS sensor and the firmware`s been altered to suit).

What I found id the ISV won`t react quickly enough on it`s own, I use ignition advance for quick correction and the ISV for load correction. My problem may have been down to the Kent Cams as well.

I target 14.7:1 at idle and it works a treat.

 

Rob: I know that my ISV will never give the results I need if it were driven directly off the MS1 PWM without the current ramp circuit in place. The PWM resulution is just far too poor. Verified this using the ISV tester (PWM generator) I built.

Keith: Good to see you about on here again! I always intended to look at your code but could never find it on the MS forum. Could you send your source code to me? I know that a commonly used trick is to put lots of ignition advance at an RPM value below idle, to 'kick' it back up whenever it starts to fall. So am I right in thinking that you've expanded on this idea by having a second finer resolution ignition table that operates in a small region about the idle RPM?

I gather that lots of advance at idle can cause excessive CO, so have you had any issues with emissions during MOT time? I ask because on a MK2 that I just put an ABF engine into, the distributor had to be wound back really far before the bugger would get through the CO test...

Trev

 

Hello Trev, It`s on the laptop so I`ll dig it out and e-mail it to you.

I coded in an "idle mode" bit which takes over under certain conditions and allows an active dashpot to settle the idle. I`ve got a 7 point ignition target table based on rpm which is active with the idle mode, this funnels the engine rpm to the set speed.

I also coded an Idle VE number which changes based on o2 input when idleing, like an active fuel trim. Basically there`s no concessions to idle in the main tuning tables, its all done seperately.

It may seem a bit over the top but I got hacked off with the poor idle from the stock code and had a go myself, KeithG on the forums did a re-write of the idle code that went into later Extra versions but I can`t see mine being included as it`s so far removed from the original.

It got through the emissions ok but the car`s an `89 and you can get away with quite a lot (3.5% max co?).

I had my codes hosted on aol but the seem to have deleted them all for me..

 

/\/\/\/\
Good skills.

in MS2 there is no closed loop idle control at all for some reason.

 

Just digging this thread up from the past...

I'm wondering if anyone out there has tried out my little expansion board. I did post this on the MS-Extra forum and I am actually quite surprised at the apparent lack of interest. Maybe it's because most people have been able to achieve good enough idle, even when PAS is operated, without ISV assistance? Maybe it's because people are using superior, higher resolution ISV control from MS2?

Still curious about whether Toyotec has had a chance to try the circuit. Ed, if time is the problem then I could assemble a board for you.

 

That would help Trev.

 

This makes me wonder whether i would need some kind of idle control with my project, considering i will be running AC also??

 

Thats a nice little circuit there Trev. Will see how I get on with my install.

One thing though, whats stopping me from using the VAG KR ISV controller box to independantly control the idle as it does now and tune the Megasquirt as is?

Gurds

 

I'm using the standard ABF idle valve. seems to work fine. Its not on closed loop, but on open loop so it opens up as it gets warmer. This is how the digifant system works, and various systems used on carb cars are probably even more crude.

Unless you have air con, big sound system etc i dont see the point in going closed loop, just adding complication thats not needed.

Mike

 

Gurds - I bet the standard KR idle controller most probably could be used just fine together with MS. I almost tried this but I was interested in getting the ISV working with MS.

This circuit has been modelled on the ISV drive circuit that's inside Digifant I and II. It is hopefully useful to people who are using MS1 firmware, because it addresses the poor ISV DC resolution that MS1 provides. It's applicable for both open-loop and closed-loop configurations. I'm actually driving the ISV in open-loop mode myself, although if necessary I may switch to closed-loop when I start running the engine with AC.

Mikey, are you using MegaSquirt? If you're driving the ISV straight from MS then presumably it's MS2?

I assume people are using MS2 these days so perhaps this circuit is quite obsolete now!

 

it comes straight from MS1, i have:

# A High Powered Idle Valve Output Transistor, capable of supplying more than 3 Amps, giving it the ability to directly drive idle control valves, such as the 3 or 2 wired Bosch units or the Ford items, etc. This means you do NOT need to purchase a Relay Board for a standard install!! The original MS idle output transistor simply isn't up to the job of controlling anything other than a relay. The ECU's I supply can directly control the idle valve! See HERE for more info.

as per http://www.extraefi.co.uk/products.htm#standard

 

Okay, so that's just a large switching transistor, so you're voltage PWM'ing the ISV directly from MS1. Which is what I didn't have any luck doing. I simply wasn't able to finely control the RPM this way, because each DC step gave a valve opening step that was much too large. Edit: Only way I could get it to work like this was to use a frequency below 100Hz, which was rather noisy; or by restricting the air to the ISV.

What frequency do you run the valve at, and what PWM duty values do you swing between when going from cold to fully warm?