SwedeSpeed - Volvo Performance Forum banner

1 - 20 of 30 Posts

·
Registered
Joined
·
174 Posts
Discussion Starter · #1 · (Edited)
I recently hacked the SUM. The operation was a success, but the patient's improvement came with too many complications/issues from related errors/shutdowns.

The alternative is to limit the max PWM duty cycle after the signal has left the SUM. So now I am trying to come up with dummy loads for the SUM, so that not having the dampers connected directly to the SUM will not cause the SUM to err out and shut down. Last night I succeeded. :cool:

I used a 5 ohm 25W wirewound ceramic resistor. WARNING: ITS POWER HANDLING CAPACITY IS MARGINAL, AT BEST.

In series with the resistor I connected an inductor:



each of these weighs 4~5 oz:


I was surprised and disappointed to see that the SUM's output changed drastically; where the 'relaxed'/base duty cycle in Comfort mode in my caR is typically around 16%, switching over to the 5-ohm-plus-coil dummy load caused the SUM's duty cycle to be much higher. Each of the 2 coils with similar labeling raised the duty cycle to 33%. The slightly smaller coil yielded 40%. If I connected the two "33%" coils in parallel (in series with the resistor), the SUM erred out. If I put the 40 in parallel with a 33, I could get a somewhat jittery 28%. Output to the dummy load in Sport was proportionally higher; Advanced was higher still; with some coils I saw percentages in the 80s and 90s. (This means that the 25W resistor is seeing roughly twice as much power as I'd figured, oops.)

So this early attempt at coming up with a SUM dummy load seems sufficient to keep the SUM from erring out (at least for a ride around the block) when it's not connected to a 4C damper, but insufficient to allow me to intercept an unaltered SUM PWM output signal and filter/modify it before passing it along to the dampers... :(


-- 100th post --
 

·
Registered
Joined
·
8,055 Posts
Most scientific way to do this would be to measure the signal to determine exactly what the inductance of the 4C coils are, and then create/purchase the right coil so it will respond exactly like the shock solenoid would.
 

·
Registered
Joined
·
547 Posts
Inductors in parallel and series follow the same laws that resistors do.

But, thanks for reminding me, I've been meaning to order a LRC meter for a long time so I can do exactly what stealthy mentioned
 

·
Registered
Joined
·
438 Posts
I may try this approach:

http://www.amazon.com/PAC-BB-6PR-Pa...rds=4+ohm+tweeter&refinements=p_85:2470955011

It pretty much has to be a prepackaged inductor/resistor. 50w capacity. 4ohm impedance.

I have a LCR meter as well. I'm going to measure the inductance today or tomorrow if I can.

4C - what may be happening is the feedback loop isn't closing appropriately causing the PWM signal to go high.

So for example - SUM, sends a 15% DC signal to the damper, and the damper damps X amount. Then the accelerometer measures the damping and says "we are good". If damping doesn't happen right, SUM sends more DC. Until it makes it to the ceiling.
 

·
Registered
Joined
·
1,186 Posts
I think 04 only can be "tricked" with resistors. 05+ is more complex, it is looking for a certain inductance. Here is what I posted about a year ago:

"
Ok, so I spent a considerable amount of time trying to replicate the inductors in the struts. I purchased an LCR meter, and some inductors off of ebay, and tried constructing several circuits to fool the 4C. The magic number seems to be 5...as in, the struts have a resistance of 5 ohms and an inductance of about 5 millihenries. Unfortunately, the circuit-board-sized inductors off of eBay had too much resistance, and I got tired of experimenting, so I scavenged from some dead front struts and built the inductors like doughy did.
"

From this thread:

http://forums.swedespeed.com/showth...-HD-BILSTEIN/page5&highlight=ohms+bigfieroman
 

·
Registered
Joined
·
3,157 Posts
I may try this approach:

http://www.amazon.com/PAC-BB-6PR-Pa...rds=4+ohm+tweeter&refinements=p_85:2470955011

It pretty much has to be a prepackaged inductor/resistor. 50w capacity. 4ohm impedance.

I have a LCR meter as well. I'm going to measure the inductance today or tomorrow if I can.

4C - what may be happening is the feedback loop isn't closing appropriately causing the PWM signal to go high.

So for example - SUM, sends a 15% DC signal to the damper, and the damper damps X amount. Then the accelerometer measures the damping and says "we are good". If damping doesn't happen right, SUM sends more DC. Until it makes it to the ceiling.
Since they are bass blockers (i.e. high pass filters), I would think they would be capacitors, not inductors.
 

·
Registered
Joined
·
438 Posts
Did my homework more on those. You are right.


Sent from my iPhone using Tapatalk
 

·
Registered
Joined
·
438 Posts
What I find bizarre is this is essentially a steel core inductor according to what I have seen. It's just a coil. Where is the 5 ohms coming from I don't get it. Most coils of this spec look to have half ohm values.

If one of my shocks is blown I'll pull the coil and see if I can find its gauge and length of conductor for our edification.


Sent from my iPhone using Tapatalk
 

·
Registered
Joined
·
438 Posts
DC Resistance is 4.9 Ohms, L = 7.5mH @1khz. Impedance at 1k is 29 Ohms.

The SUM module is powered by a 15A fuse - assuming 14v, this means that the module and all 4 shocks must share max wattage of 210W. its unlikely the control module only uses 10W, but possible. So call it 50W per shock possible power output.

If it were a 5Ohm impedance at operating frequency (I don't know that it is, I think it is actually 75 ohms of impedance based on the LCR), then each shock should be rocking 20W or so.

Reading it seems that VIDA says the outputs are 0.5A current draw, So that would indicate 7W. 7W would indicate somewhere in the 29 ohm range.
 

·
Registered
Joined
·
174 Posts
Discussion Starter · #10 · (Edited)
This post combines sprecific reply info with general info that the respondee and some readers very likely already know, but some may find enlightening.

4C - what may be happening is the feedback loop isn't closing appropriately causing the PWM signal to go high.

So for example - SUM, sends a 15% DC signal to the damper, and the damper damps X amount. Then the accelerometer measures the damping and says "we are good". If damping doesn't happen right, SUM sends more DC. Until it makes it to the ceiling.
That was my working hypothesis for years. After all, as the SUM adapts following a recalibration, it surely avails itself of body accelerometer data (IOW, it checks [the results of] its work). I was imprisoned by this belief; for years it kept me from even thinking seriously about modifying the SUM's output signals. In part (it was fueled) because no one had successfully 'fooled' the SUM using dummy loads. Jan Hargis (sp?) had some success, but, iirc, his resistance-only experiment failed on '05+ Rs. (Plus I'm not even sure it was all 4 corners anyway...) Then dougy set his mind to doing away with 4C and with the resulting error; he hacked apart his old dampers, salvaged the coils, plugged them in, and (once he included ferrous material inside the torroid) never looked back! That is what, finally, earlier this year, freed me from my former working hypothesis...

OTOH, both views could be correct; just because we now know that it is possible to keep a SUM with ineffectual output from setting errors does not necessarily mean that the SUM won't be so confused by its lack of effectiveness that its signals won't become unusable.

Since they are bass blockers (i.e. high pass filters), I would think they would be capacitors, not inductors.
good thinking



When the PWM signal goes from high to low (could be zero, but...), it (ideally) is a vertical line with no slope or overshoot/ringing. The signal is ideal within the SUM. But once it gets applied to, and applies power to, an actual load, deviations from ideal are inevitable.

When a current flows through an inductor, a magnetic field is generated. When the flow stops, the field collapses, which attempts to cause a reverse flow. (This is commonly referred to as a back-EMF. EMF is a fancy way of saying voltage. It is why it is not uncommon to see a diode connected across the terminals of a solenoid...)

So when the PWM signal that's been going through an inductor momentarily ceases, and the voltage falls to 'zero', it actually very briefly falls to below zero. IOW, as it falls, it overshoots zero.

OTOH, a capacitor stores electricity. So a capacitive load will let the trace down gently, kind of the way a non-clenched shock aborber lets the car down gradually/gently. So that trace starts to fall vertically but then levels out before/as it approaches zero.

If you can visualize this, you can see that, in some sense(s), an inductor is the opposite of a capacitor.

The SUM not only visualizes this, it more or less 'tracks' it. The SUM has a fairly high precision ammeter built into each output circuit.

What I find bizarre is this is essentially a steel core inductor according to what I have seen. It's just a coil. Where is the 5 ohms coming from I don't get it. Most coils of this spec look to have half ohm values.
I share your puzzlement. My current working hypothesis is that the wire is an imperfect conductor. Placing resistance in series with an inductor makes the inductance less than ideal... anyway, the ~ 5 ohm resistance must be "built in" (to the windings).

DC Resistance is 4.9 Ohms, L = 7.5mH @1khz. Impedance at 1k is 29 Ohms.
Thanks for adding your voice/info. I've seen DC resistances between 4.7 and 4.9 ohm (at the SUM). I've seen inductances (published heRe) between 5 and 6.34 mH. My experiment was very scientific (I found that the dummy load altered the output signal, and that it was not possible, using this approach/resistance, to lower the inductance enough to not alter the signal's duty cycle. BTW, this was happening immediately, in real-time; I still have no reason to believe that this alteration was any kind of adaptive response to the damper's lack of responsiveness/effect perceived via body accelerometer. No one else who's used [the original damper coil as] a dummy load has any idea -- or probably any need to care -- if their load alters the SUM output signal or not), just not quantitative.

The PWM freq is 2.35 KHz. Of course, that's not a sine wave, just how often the abrupt ON/OFFs occur.

The SUM module is powered by a 15A fuse - assuming 14v, this means that the module and all 4 shocks must share max wattage of 210W. its unlikely the control module only uses 10W, but possible. So call it 50W per shock possible power output.
That agrees with simply looking at the power consumed by presenting a 5 ohm load to a 15V source. P = V * I. We know from Ohm's law that I = V / R = 15 / 5 = 3A, and also we know from Ohm's law that V = I * R, so P = I * I * R = 3 * 3 * 5 = 45W

One of the reasons I'm disappointed that I'll need to come up with my own output drivers in order to introduce a PWM duty cycle limiter is that the SUM is fancier, less straightforward, that most people, myself included, ever dared imagine. The PWM seen by the dampers varies between 14.4V (peak) and -1.8V (peak). Even allowing for the negative overshoot peak, its 'zero' is still somewhere between -1.2V and -1.5V. (This negative offset is sufficient to cause one of my meters to see the average output voltage for a 20% duty cycle as 2V; IOW it 'thinks' the PWM is between 0 and 10 volts.)

If it were a 5Ohm impedance at operating frequency (I don't know that it is, I think it is actually 75 ohms of impedance based on the LCR), then each shock should be rocking 20W or so.
Keep in mind that if the duty cycle becomes doubled by the passive dummy load, your estimate could be off by half. And those impedance numbers sound high to me. OTOH, if the damper coil really does present that much impedance, then they're really not handling anywhere as much power as (admittedly DC resistance) theory and fuse size (which probably allows for 100% duty cycle, and thus resistance not impedance) both predict.

Hmmmmm..... I guess my next attempt will be with higher resistance ... or maybe measure things with an old REMOVED actual damper/coil ... I did no current measurements, but I know the resistor warmth/heat I felt was real, and the dummy load was close enough to keep the SUM from erring out.

When the coil is removed from a 4C damper, does anyone know if the metal shaft which protruded into its center moves up and down freely (and how far/easily)?
 

·
Registered
Joined
·
438 Posts
It probably must move - the central part. It would be the active component in the mechanical damper, and the only reason it should move.

Did you happen to hook an ammeter into the circuit to determine the load during operation? I have an ammeter that may do if needed, but if we have this amperes measurement it would be useful.

A note on your resistor - if it is 5Ohms, it will dissipate 39.2W at full duty cycle. So, its going to soak up 2.8A of current at 100%DC, but since it never runs at that it would only be 40 something percent of that.

Still, it would be rocking quite a bit of load.
 

·
Registered
Joined
·
438 Posts
Actually, looking at this more, the 5Ohms of DC resistance serves no purpose to the operation of the shocks internal mechanism.

Its built into the windings to act as a protection for the output driver of the SUM. That way if the shock is damaged by impact/etc (especially plausible in the rear shock absorber), the resistance can never be compromised past the 5Ohm value of the winding wire. In a regular inductor the base current could become EXTREMELY high during a coil damage scenario. Since this is a PWM signal and not a true AC sine wave, the negative EMF doesn't occur for the most part. You would need to protect against a base DC current that way. It also sets a floor for energy usage by the damper at 100% Vin, since the field is no longer collapsing causing an inductive current flux and the bloody thing just acts like a length of wire. If it were low Ohms, it would burn up immediately.

Translation - the 1KHZ response of this inductor at 29 Ohms of impedance is actually 5 ohms of pure resistance and 24Ohms of inductive reactance. So the inductor should be 4mH, and the resistor should be 5Ohms.

Its impossible to calculate the equivalent resistance though - we don't exactly know yet what the amps are with a PWM for the advanced setting. It may be measurable. If it's not an inductor though, it won't vary by duty cycle, and the SUM may error.

I will see if the SUM is stupid enough to listen to a 75Ohm resistance which is close to a 4mH inductor in series with a 5 ohm resistor operated at 2.5khz. The PWM will approximate a SIN wave in some ways, but I just can't tell which ways. It has a higher peak voltage value but no interstitial voltage and practically no negative voltage.

Side note, the negative voltage you see shouldn't be coming from the driver of the SUM - it should be the inductive field collapsing.
 

·
Registered
Joined
·
174 Posts
Discussion Starter · #13 ·
It probably must move - the central part. It would be the active component in the mechanical damper, and the only reason it should move.

A note on your resistor - if it is 5Ohms, it will dissipate 39.2W at full duty cycle. So, its going to soak up 2.8A of current at 100%DC, but since it never runs at that it would only be 40 something percent of that.
I can visualize the theory just fine, I was hoping to hear from one of the folks who've actually had one cut open, and fingered it. It's probably minimally (though firm enough to be moved/controlled by coil current but not by vertical accerations from bumps themselves, right?) spring-loaded with a funny shape at the other/inner/hidden/working end, to provide such firm damping with 0% duty cycle in its firm limp(/stationary/off) mode.

Just keep in mind that the inductance varies depending upon whether or not the coil is fre-air or has a ferrous core, as dougy and his SUM have written/demonstrated. And if that core moves, and is not always fully within the coil, the inductance will vary... So it's possible that no passive dummy load can perfectly mimic the actual load. Fortunately, my experiment demonstrates that an imperfect mimic can suffice, especially if one is disinterested in the actual duty cycle of the PWM signal going through the dummy load.

15V / 50 ohms = 0.3A; .3 * .3 * 50 = 4.5W at full duty cycle

IOW, down around 5 ohms, a maximum of roughly 50W is dissipated. At higher resistances, less current flows, less power gets consumed...
 

·
Registered
Joined
·
1,090 Posts
honest question, what is the point of this? You can just cut out the inducers from your old 4C shocks, plug them in, and done! no more dash messages!
 

·
Registered
Joined
·
438 Posts
Sorry for my ramblings - I have a degree in EE but haven't used it in 26 years so its rusty.
 

·
Registered
Joined
·
438 Posts
The point of this is that I have 2 and perhaps 3 of my 4c shocks in good shape. They are newer replacements. I'm going to coil overs. I would like to give the folks here a shot at some lower mileage shocks instead of just butchering my way to a solution LOL.
 

·
Registered
Joined
·
174 Posts
Discussion Starter · #17 ·
The SUM is powerful and sophisticated. It is immune to coil damage; the SUM outputs are fully protected against complete short circuit.

If you re-read what I wrote, you should understand that the instantaneous small negative overshoot is from the coil's back-EMF, but the steady negative voltage is from (driven/acused by) the SUM.
 

·
Registered
Joined
·
174 Posts
Discussion Starter · #18 ·
You can just cut out the inducers from your old 4C shocks, plug them in, and done! no more dash messages!
1. "just"? haveyou done that? do you have any idea how much work that is and how long it takes and what risk is involved?

2. dougy, the first to successfully fool neweR 4C with dummy loads, did that and it failed; he had to increase the inductance...

3. different strokes for different folks..
 

·
Registered
Joined
·
438 Posts
Interesting on the negative voltage - it may be the clamping force used to hold the dampers post in an out-of-actuation position.
 

·
Registered
Joined
·
174 Posts
Discussion Starter · #20 · (Edited)
Here is a look at 3 (C/S/A) settings' SUM damper output. (This freq counter says 2.35 on the logic signals but 2.32 on the outputs, even though they're in 'perfect' sync.):



I guess I overstated the steady negative voltage (maybe I remembered spacing with respect to the trigger cursor rather than the trace/0 cursor?), However, you can see that it does not sit at zero at rest, but nearly 1 volt negative. When I tried various different inductors looking for a viable dummy load, not only did I see larger/smaller negative overshoots, but also more tiny blips (as are evident in 4 of the 6 low-horizontal segments). Maybe the SUM does tiny, functionally insignificant, load-testing stuff during the off portion of the duty cycle?
 
1 - 20 of 30 Posts
Top