Wildguzzi.com
General Category => General Discussion => Topic started by: PJPR01 on September 13, 2019, 11:53:00 PM
-
So...I'm going thru my annual GuzziDiag checkup on the Norge, doing the Throttle body balance, and checking a few parameters.
GuzziDiag tells me my voltage is 12.8 or 12.9 at idle, barely moves at 3k rpm from that setting, occasionally blips up to 13.0, but doesn't stay there.
I decide to check the voltage using another source by plugging in a high quality device into my cigarette lighter socket near the battery, and it also reads 12.8/12.9 at idle.
I pull out my test meter and attach it to positive/negative terminals and it reads 13.2 to 13.3 while the bike is running.
So which one am I to believe now? Anyone else seen this discrepancy with GuzziDiag *(version V0.47), or do other folks see the same reading on their Norges?
Thanks!
-
I used to see a similar voltage drop between the battery terminals and the dash readout on my B11.
-
Isn't that voltage low, no matter how it's tested? I would think around 14V would be "normal" at cruising rpm.
-
As already written:
The ECU measures the voltage not at the battery poles.
There is at least one fuse and several contacts between.
This results in a lower voltage which is seen by the ECU.
There is a correction factor available, but ... don't try to adjust it.
Since there are many values which depends on the voltage, like injection and ignition.
Less voltage means more time is needed for the coils and the injectors.
If you adjust the voltage value to be more real, the engine runs verybad, until you also correct the voltage dependend tables.
To much todo for only getting the 'real' voltage.
But that's alos a reason why a Guzzi is not starting, even if you measure 12.0V at the battery.
The ECU measures 11.5 V and don't allow to start.
-
Use your multimeter to measure the battery Voltage, you can connect directly to the battery with that.
Guzidiag is telling you the Voltage at the ECU, i'm guessing after the Kill Switch.
-
Use your multimeter to measure the battery Voltage, you can connect directly to the battery with that.
Guzidiag is telling you the Voltage at the ECU, i'm guessing after the Kill Switch.
I thought that was what he was saying..
I pull out my test meter and attach it to positive/negative terminals and it reads 13.2 to 13.3 while the bike is running.
-
My G8 has never shown more than 13.4 on the dash. Always measures just under 14 at the battery. Never fails to start, if I haven't ridden for a week or more in winter the dash may drop out before it fires but it always fires up. Never suffered from Startus Interuptus either.
Stelvio on the other hand always reads 14.2, never drops the dash out, always has 14.2 at the battery, also always starts!
(Shrug?)
Having read Beard's explanation I am going to immediately cease giving it any more thought, not that I was exactly straining my brain over it before this! :grin:
Pete
-
Your battery is old and tired
-
Mine? Ummm? No.
-
Pete
I presume by G8 you mean Griso 8 Valve
If it's wired this way it will eventually suffer
Perhaps the factory wired the bikes for OZ better.
http://www.thisoldtractor.com/guzzi007/schematics/2008_Griso_8V.gif
The Start relay is powered from a switched fuse
The fix is to cut the yellow wire feeding start relay and wire it from a direct supply
The Earlier 2 Valves were wired direct but then they used a tiny wire from relay to solenoid this one has a second Startup Maintenance Relay, a waste of space IMHO
Out of curiosity watch the number plate light while someone else starts the bike
-
Thanks everyone...will check a few scenarios now including the battery...perhaps it is fatigued, charging it again today to see if it's holding voltage or dropping.
However, assuming the battery does check out to be fine, how to tell whether it’s the alternator, alternator belt or voltage regulator that is the potential culprit for the low voltage output.
Presumably, change the belt, test it again and see if there has been any improvement would tell me but is there another way to know?
Thanks all!
Just for fun, I'm going to run the same tests on the Scura and Griso and see what sort of readings I get there.
-
If the belt is 5 years old or has 50,000 miles, replace it on general principles any time you are near it. More often if you live in the desert...
AMHIK :grin:
-
Good point... It's due for changing...it's 10 years old now.
Good thing is that now with the battery freshly charged and holding , the voltage is reading 13.8 when idling, so we're back up into good territory. I don't think I've ever seen it higher in the past.
-
13v max at the battery shows a problem which will also give you poor lighting.
I had exactly this on my Norge. I wondered why I couldn't see the road ahead.
I traced it to a burned connector hidden in the loom on the main output wire from the alternator to the battery (via that 40A fuse). That connector is roughly above the relays on the RH side.
I added another wire in parallel to the original back to the battery and immediately the volts went up to just over 14, and the dash voltage was also reporting 14v.
I can also now see where I'm going.
-
Old battery or new, regardless of the method, my dash has always shown a lower voltage than any other method of measurement; GuzziDiag and an appropriately applied meter have generally been close to each other and a bit higher than the dash, but nowhere near what I would expect (13.2 or so vs the 14 I see on my 8V Guzzis).
Belt quality is not a factor for me - this occurs consistently.
-
The dash measures the Voltage on the yellow wire after the ignition switch which also happens to be the same weak supply that causes Startus Interuptus
Guzzi Diag I assume reads the Voltage from the ECU via a different route or does it also read it from the dash?
Meinolf pointed out that its the ECU that measures the Voltage and sends it to the dash via Canbus.
I think the ECU gets power through the main Injection Relay (29) via pin 17 which is much more direct than the dash.
Lets assume PJPR01 hasn't done the Startus Interuptus fix yet, somewhere he is losing about 1 Volt.
http://www.thisoldtractor.com/guzzi007/schematics/2007_Norge.gif
Under normal conditions the Voltage on the yellow wire should be very close to the battery but the instant you press the start button it will be lower for example lets assume the battery is 13 Volts
Suppose there is 0.2 Ohms resistance between the battery and the yellow wire where it goes to the dash
The start solenoid is pulling 30 Amps, 30 x 0.2 = 6 Volts leaves only 7 Volts at the dash, is it any wonder it turns up its toes?
Meinolf pointed out that its the ECU that measures the Voltage and sends it to the dash via Canbus.
Once the starter is engaged the solenoid drops down to 10 Amps but that's still 10 x 0.2 = 2 Volts leaving 13 - 2 or 11 for the dash
(0.2 Ohms is extreme but as the resistance drops the solenoid inrush current goes up, 50 Amps with a full 12 Volts)
After doing the Startus Interuptus fix the solenoid current is no longer through the yellow wire its bypassed by a new fuse so the current through the original circuit is lets say 1 Amp
1 x 0.2 = 0.2 Volts that leaves 12.8 at the dash
The only true way to read the battery Voltage is with a Voltmeter at the battery terminals, anything else is just an approximation.
-
Hi Roy,
please allow me to add my 10 cents.
The dash measures...
I'm sure the dashboard monitors the voltage, lot's of power hungry components are activated through the dash, but the voltage shown in the display is coming from the ECU in a CAN packet.
Guzzi Diag I assume reads the Voltage from the ECU via a different route or does it also read it from the dash?
Yes, GD shows the voltage as read by the ECU via one of the several ADC ports. As Beard pointed out there's a voltage calibration factor used in the code to correct the analog voltage to the proper digital value. Any discrepancy between the voltage shown in the dashboard and the one shown in GD "can" be corrected with the calibration factor, but this will influence any calculation in which the voltage is used. Such as injector trim. Changing the calibration factor to sync voltage shown between dash and GD will accordingly impact injector pulse duration, impacting fuel injected. I did change the calibration factor some time ago to fix a 0.4V discrepancy, the result was a large increase in injected fuel across the board. I fully support Beards advice to leave this alone and live with discrepancy between the two values shown.
The only true way to read the battery Voltage is with a Voltmeter at the battery terminals, anything else is just an approximation.
Full agreement
Cheers
Meinolf
-
Thankyou Meinolf
There you have it guys
The Voltage drop on the yellow wire should not effect the dash although if it goes too low the solenoid will not pull in (Startus Interuptus)
It won't show up on the voltage display because battery Voltage is read by the ECU on a different circuit, a more direct connection to the battery.
If the battery Voltage is too low the ECU will disable the starting circuit (this is not classic Startus Interuptus)
A couple of questions Meinolf
You say the Voltage is measured via "one of the several ADC ports",
Do you know which one (pin no?) or is it 17 also
I assume 17 supplies the ECU loads(supply for discrete outputs is it also an ADC?
-
Hi Roy,
You say the Voltage is measured via "one of the several ADC ports",
Do you know which one (pin no?) or is it 17 also
I assume 17 supplies the ECU loads(supply for discrete outputs is it also an ADC?
the usage of the ECU pins changes between the BINs 2230 and 3222. Below is for the 2230.
in adc1 (index 47) ECU pin V23 analogue input (not connected)
in adc2 (index 48) ECU pin E13 air pressure sensor (not connected)
in adc3 (index 49) ECU pin E03 TPS
in adc4 (index 50) ECU pin V04 battery voltage*0.2
in adc5 (index 51) ECU pin E05 engine temp sensor
in adc7 (index 52) ECU pin V22 lambda probe
in adc9 (index 53) ECU pin V31 right lambda probe (not connected)
in adc11 (index 54) ECU pin V11 monitor*0.68 lambda heater
in adc12 (index 55) ECU pin V01 monitor*0.68 starter relay
in adc13 (index 56) ECU pin V14 monitor*0.68 not connected
in adc14 (index 57) fixed voltage reference
Cheers
Meinolf
-
Thats very interesting Meinolf
http://www.thisoldtractor.com/guzzi007/schematics/2007_Norge.gif
Looking at the diagram and your list it seems like the top connector pin 4 is measuring the battery Voltage, tracing the wiring it has a Predisposition Condenser and a Predisposition Diode listed.
I don't know what i'm to make of those two items, it almost looks like something tried in the past to capture the peak battery Voltage.
Am I right in thinking the capacitor doesn't exist and the diode is jumpered out?
That would surely make a big difference to the Voltage reading? 0.6 higher than with the diode.
Removing the diode would require a different injector time offset would it not?
Anyway it looks like the ECU has a dedicated fuse "E" to measure the battery Voltage directly so it shouldn't be effected by the drop through the yellow wire so it should read almost identical to a meter across the battery :thumb:
-
I found what I think is the connector for the diode, and there is just a loop of wire on my Norge.
-
Hi Roy
..., tracing the wiring it has a Predisposition Condenser and a Predisposition Diode listed. I don't know what i'm to make of those two items, it almost looks like something tried in the past to capture the peak battery Voltage.
the diode and capacitor/condenser are part of a kit from MG for NORGE 1200, 850 BREVA 1200, 1100, 850, 1200 SPORT 2V having start issues with a cold engine. The procedure is described in this service bulleting (french, I'm afraid) https://www.fichier-pdf.fr/2012/09/02/stage-guzzi/preview/page/56/ https://www.fichier-pdf.fr/2012/09/02/stage-guzzi/preview/page/57/
Specs of the parts: AP9100739 electrolytic capacitor: min. 47.000 µF / >= 63 V, LowESR and extended temperature range up to 85°C
Diode: AP8212588 Schottky 5A/200V for example MBR10200CT-LJ (2 diodes, need to be connected in parallel)
Am I right in thinking the capacitor doesn't exist and the diode is jumpered out? That would surely make a big difference to the Voltage reading? 0.6 higher than with the diode. Removing the diode would require a different injector time offset would it not?
AFAIK the diode has no impact on the voltage reading, it serves as protection for the capacitor. Certainly no influence on the injector timing.
Cheers
Meinolf
-
The diode even a Shottky must have some effect on the Voltage reading due to it's forward bias for example if the battery is at 12.5V the ECU will see 12.5 - 150 to 450mV
The injector is fed the Voltage direct through the Injection relay and the other side is grounded via a transistor in the ECU
The Schottky diode (named after the German physicist Walter H. Schottky), also known as Schottky barrier diode or hot-carrier diode, is a semiconductor diode formed by the junction of a semiconductor with a metal. It has a low forward voltage drop and a very fast switching action. The cat's-whisker detectors used in the early days of wireless and metal rectifiers used in early power applications can be considered primitive Schottky diodes.
When sufficient forward voltage is applied, a current flows in the forward direction. A silicon diode has a typical forward voltage of 600–700 mV, while the Schottky's forward voltage is 150–450 mV. This lower forward voltage requirement allows higher switching speeds and better system efficiency.
While adding the diode would not change the injector timing directly it would change the Voltage the ECU reads and therefore it will change the digital readout on the dash compared to a digital meter applied directly to the battery it must also change the Map in some way.
A 47µF capacitor is quite significant, it probably takes the ECU Voltage input a few seconds to discharge that under cranking so it will make the ECU think the battery is a little higher than reality, does it just have to fool the logic?
I'm sure that the factory could have fixed the starting much better by making a change in ECU logic rather than trying to fudge it. On other models they added a Startup Maintenance Relay to override the ECU controlled Start Relay.
Have you ever seen a document describing the ECU logic? It bugs me that I can make changes in the controls of a chemical plant but I'm not able to understand how my simple motorcycle is programmed.
-
Hi Roy,
The diode even a Shottky must have some effect on the Voltage reading due to it's forward bias for example if the battery is at 12.5V the ECU will see 12.5 - 150 to 450mV. While adding the diode would not change the injector timing directly it would change the Voltage the ECU reads and therefore it will change the digital readout on the dash compared to a digital meter applied directly to the battery it must also change the Map in some way.
your reasoning is sound and I don't disagree. The fact of the matter is that I've never measured the effect of the diode included in the Guzzi fix on injection times. I did measure the significant effect of changing the voltage calibration factor on injection time, however. Apart from measuring, it is also directly noticed when riding the bike. I've not heard of such effects as a result of a diode installed.
A 47µF capacitor is quite significant, it probably takes the ECU Voltage input a few seconds to discharge that under cranking so it will make the ECU think the battery is a little higher than reality, does it just have to fool the logic? I'm sure that the factory could have fixed the starting much better by making a change in ECU logic rather than trying to fudge it. On other models they added a Startup Maintenance Relay to override the ECU controlled Start Relay.
Well, the capacitor buffers the voltage during starter engagement and avoids the low voltage error. Basically a different approach than changing the wiring to avoid the startus interruptus. And yes, changing the respective parts in the code would have achieved the same result. Your guess why this wasn't done is as good as mine.
Have you ever seen a document describing the ECU logic? It bugs me that I can make changes in the controls of a chemical plant but I'm not able to understand how my simple motorcycle is programmed.
No, I haven't seen any document from Marelli or Guzzi describing the ECU logic. However, a small team consisting of John Th. from the UK, Beard and myself, have been reverse engineering the 5AM, focused on the 2230, code beginning 2 years ago. The result is a good, if not yet complete, understanding of the code. The library approach used in the code also served well to reverse engineer the BINs used in Ducatis and other makers. I've posted the findings on this (https://wildguzzi.com/forum/index.php?topic=93758.90) and other boards and invited anybody interested in participating in the enterprise to join. Nobody did....
Cheers
Meinolf
-
So it's just a simple way of boosting the voltage in cold weather when the battery volts drop too much.
Not dissimilar to the Ultracapacitors used in some cars with stop start technology.
The diode is there to stop current flowing back towards the battery.
-
Hi Roy,
your reasoning is sound and I don't disagree. The fact of the matter is that I've never measured the effect of the diode included in the Guzzi fix on injection times. I did measure the significant effect of changing the voltage calibration factor on injection time, however. Apart from measuring, it is also directly noticed when riding the bike. I've not heard of such effects as a result of a diode installed.
Well, the capacitor buffers the voltage during starter engagement and avoids the low voltage error. Basically a different approach than changing the wiring to avoid the startus interruptus. And yes, changing the respective parts in the code would have achieved the same result. Your guess why this wasn't done is as good as mine.
No, I haven't seen any document from Marelli or Guzzi describing the ECU logic. However, a small team consisting of John Th. from the UK, Beard and myself, have been reverse engineering the 5AM, focused on the 2230, code beginning 2 years ago. The result is a good, if not yet complete, understanding of the code. The library approach used in the code also served well to reverse engineer the BINs used in Ducatis and other makers. I've posted the findings on this (https://wildguzzi.com/forum/index.php?topic=93758.90) and other boards and invited anybody interested in participating in the enterprise to join. Nobody did....
Cheers
Meinolf
Hi Meinolf
Well I think we both agree that changing the Voltage effects the injection time, the ECU adds time as the Voltage drops. Adding a diode in series with an electrical load is a common method of lowering the Voltage on the load because of the Voltage lost in forward biasing the diode. Guzzi used a Shottky diode because the forward bias is a lot less than a Silicon 0.2 - 0.3 Volts rather than 0.6.
They added a large capacitor to hold the Voltage at the ADC artificially higher and the diode prevents the capacitor from draining back into the battery when the starter current spike pulls the battery down.
The thing is they don't seem to understand why the Voltage takes a spike down, the inrush to the starter solenoid can be over 50 Amps and thats before the starter even turns over pulling 150+ Amps, they don't show the solenoid correctly on any schematic, I really believe they don't understand it. Anyway they must have since scrapped the diode / capacitor at the schematic I attached shows them missing. At that magnitude of currentit only takes 0.01 Ohms to drop half a Volt
There is really no reason why the ECU reading of the battery Voltage (and assuming Guzzidiag uses that same ADC value) shouldn't be exactly the same as putting a multimeter across the battery. The Shottky diode would explain why the former would read lower), the supply to pin V04 must be only a few milliamps so there would be little Voltage drop.
I suspect there may be very few bikes with the diode still in place.
As to the logic, I think there must be some logic that says if the battery Voltage is too low disable the start relay. On some of the CARC bikes they added a Startup Maintenance relay, that latches in the start circuit bypassing the Start relay controlled by the ECU logic.
I have studied enough Guzzi schematics to see a pattern in their thinking, they have been trying to overcome Startus Interuptus in various ways for at least 20 years when really the cause is so simple I maintain they think the solenoid only draws 10 Amps they choke off the current the solenoid needs.
-
So it's just a simple way of boosting the voltage in cold weather when the battery volts drop too much.
Not dissimilar to the Ultracapacitors used in some cars with stop start technology.
The diode is there to stop current flowing back towards the battery.
Yes, boosting the apparent Voltage
Actually it doesn't boost the Voltage, just captures the peak Voltage for a few seconds, its a basic electronic trick, not very elegant.
An Ultracapacitor would store some significant energy (commonly called Supercapacitors)
The 47 microfarad capacitor is quite large to the ECU Voltage reading Analog to Digital Converter but insignificant to any real load.
I'd be surprised if they are still fitting it, Im sure they dropped it when it didn't work.
Oh yes Brian, where did you find that connector?
The capacitor slot (62) is probably just there also, you could follow the wire from (61)
If you break that diode connection I doubt the bike would start. (it wouldn't know the battery Voltage)
I'd be interested to know how the Voltage at (62) to chassis compared to the Dash readout
Have you done a Startus Interuptus mod on your bike?
-
I did the start issue mod years ago. Did it on my 1980 V50 and on my 1990 SPIII. Then on the 2007 Norge. Guzzi never learn it seems.
Can't remember where that loop was now exactly, but under the seat not far from the relays.