Club 80-90 Forums

937carrera wrote: A few weeks late, but something I saw over on the Samba while trying to help itchy with his temp gauge was relevant here.

On the pre '85 vehicles the spec for empty is at the end of the reserve, post '85 the spec is given for the start of the reserve area, you have to look at the pictures as well as the text.

Ah, that's a useful snippet. So that makes the values slightly more consistent:

34 Ohms = Post-85 Full
37 Ohms = Pre-85 Full
168 Ohms = Post-85 Start of Reserve
288 Ohms = Pre-85 End of Reserve

bigbadbob76 wrote:I just checked my 220uF 16V electrolytic and it's 8.15mm diameter.
It might be worth having pads for a small capacitor between pin 2 and ground for ignition suppression in case of noise pickup, probably a 10nf or 22nf would be fine.

Noted, thank you.

bigbadbob76 wrote: Surface mount soldering is not too bad if you have a fine tipped iron you can do it manually.
Using physically larger sized resistors/capacitors helps, don't be tempted to use the tiny stuff at first until you get the hang of it.
A simple circuit like this is ideal for cutting your surface mount teeth on.

I have a friend at work who does a lot of surface mount. He's the one who introduced me to the PCB manufacturer in Shenzen. He orders a solder mask when he orders his PCBs, uses it to apply a solder paste to which he sticks the components, then sticks the whole lot in an oven to melt the solder. I may well have a go with this board, but I won't be doing it that way!

CJH wrote: 34 Ohms = Post-85 Full
37 Ohms = Pre-85 Full
168 Ohms = Post-85 Start of Reserve
288 Ohms = Pre-85 End of Reserve

And suddenly all becomes clear.

CJH wrote:
I have a friend at work who does a lot of surface mount. He's the one who introduced me to the PCB manufacturer in Shenzen.

I have one of those friends too, lol.
I havn't gone down the surface mount design route either but have done surface mount re-work which is not too bad with a steady hand and a mgnifier.

bigbadbob76 wrote:I just checked my 220uF 16V electrolytic and it's 8.15mm diameter.
It might be worth having pads for a small capacitor between pin 2 and ground for ignition suppression in case of noise pickup, probably a 10nf or 22nf would be fine.

I adjusted the PCB to take a bigger C1 (although the one in my RS basket is only 6.3mm), and I added a space for a 22nF multilayer ceramic capacitor - do let me know if that one (or the footprint) isn't suitable. Does it matter which side of the 8k2 resistor it goes?



I'm not ready to order yet - I need to make the most of the £25 DHL Express shipping, so I'll have a go at the surface mount version first, and maybe add the necessary components as well.

Looks good to me Chris.
That C2 should be fine and location is good.

Great, thank you.

I had a go at swapping some through-hole components for SMD equivalents. I left the connector block and variable resistor as through hole. I used fairly large packages for the resistors (1206) and C2 (0603), and C1's size is dictated by its properties (8.3mm x 8.3mm). In the end I don't think it saves a great deal of space on this particular board. And the SOP8 versions of the MC3423 that I've been able to find online seem far more expensive than the DIL8 version, so I'm tempted to switch that back to through hole.

I had been using 1mm tracks, which were far bigger than needed but will be robust for the connections to the outside world. But 1mm tracks didn't work well with the small feet of the SOP8 IC, so I've used 0.5mm tracks for all but the external connections.

I took the mk2 for a spin tonight.
With the gauge showing just below the 1/4 mark during straight and level flight, the light came on for 4-5 seconds on left hand bends and not as much during acceleration and up hills.
As expected really. The sender is front left of the tank.
I think we might need a bigger capacitor on the mk3.
470uF 16v should delay it nicely.



Sent from my iPhone using Tapatalk

bigbadbob76 wrote: I think we might need a bigger capacitor on the mk3.
470uF 16v should delay it nicely.

Sounds good - an occasional light when the fuel's getting low is no bad thing.

I think I found out why these MC3423 ICs are expensive in the SOP8 form factor - are they an obsolete chip? Is there another 'crowbar' chip that would swap in?

PS. I thought we were on Mk3 already.

Yes, probably are obsolete now, I had a few in my bottom drawer so gave it a try.
maybe best to stick with the through hole dip8 version then.

Yes, we're on Mk3 but I was testing my mk2.
I havn't done a gauge mount version of mk3 yet but my slosh findings with mk2 will apply to mk3.

Another check tonight during my other tests.
Mk2 circuit with Golf Mk1 gauge.
Parked level, led just glowing-


And 12 miles later, back in the same spot, led glowing bright-


The board was set up on the bench with my, now spare, late gauge. So this gives a good reference check of the accuracy of my golf gauge.



Sent from my iPhone using Tapatalk

I filled up recently and my gauge didn't show full after filling,
having done 52 miles it's now showing half full and my Mk2 low fuel light is coming on briefly when going round left hand bends.
That's too annoying to live with.
So I've fitted my Mk3 prototype with a 470uF delay capacitor and will see how that goes.
I'm thinking either something's upsetting the gauge reading or my Golf fuel gauge is not reading the same as my original T25 one.

Golf Mk1 sender is quoted on the net as being 39 ohms full and 287 ohms empty which ties in close enough with your posted T25 resistances.
And the light came on fully at just touching reserve (same as it did with my T25 one.)
so the volts are right for the gauge but it's reading low.
Note to self, must check sender earth.

CJH wrote: 34 Ohms = Post-85 Full
37 Ohms = Pre-85 Full
168 Ohms = Post-85 Start of Reserve
288 Ohms = Pre-85 End of Reserve

E D I T- just been to the co-op for my lunch,
Mk3 fitted, LED still coming on very briefly on left hand junctions so 1000uF capacitor fitted, that should give a 10 second trigger delay.
we might need more yet for fast left hand bends.

Ok... with 1000uF delay capacitor it doesn't trigger on hills, bends or junctions at half full.
Time will tell at lower levels.

So I did a quick gauge comparison tonight.
Sender at 89.2 ohms-
Golf gauge at 6.34V.


T25 gauge at 6.37V


Thats close enough for ministry work.

So either the pump cut off early, my sender earth has suddenly gone bad, or I'm drinking fuel like crazy.


Sent from my iPhone using Tapatalk

bigbadbob76 wrote: Mk3 fitted, LED still coming on very briefly on left hand junctions so 1000uF capacitor fitted, that should give a 10 second trigger delay.
we might need more yet for fast left hand bends.

I'm impressed that you can predict the delay for a given capacitor value. They've always been a bit of a mystery to me! An even longer delay wouldn't be a problem in practice - even minutes would be ok, because once the fuel level is genuinely low it stays that way for a relatively long time, and it's not as if the warning is a time-critical issue (unlike overheating for instance).

I haven't dared to take my new tacho-compatible fuel gauge anywhere near empty yet, as I'm not confident in the way it will work at the bottom end. I need to take some measurements and get a bit of confidence in it, and maybe carry a spare gallon when it gets close to empty. But it does read full when the tank's full though

CJH wrote:
I haven't dared to take my new tacho-compatible fuel gauge anywhere near empty yet, as I'm not confident in the way it will work at the bottom end. I need to take some measurements and get a bit of confidence in it, and maybe carry a spare gallon when it gets close to empty. But it does read full when the tank's full though

It'll be fine, and that's probably what you will need to do.

On some later VW's the fuel gauge operation is checked by draining the tank and adding 7 litres of fuel, then needle should then be aligned to the top of reserve.

The chip datasheet gives you the time formula. minutes would be good but would need a HUGE capacitor. 1000uF = 10s so 6000uF = 1 min. We have to compromise.


Sent from my iPhone using Tapatalk