USB isolator is very good idea, I ordered from my source and will report when it arrive.

having the source of the disturbances identified as the ignition components, I have the doubts at which end of the USB the comm fails, blue box or PC.

If the disturbances are induced to the sensor cables (load cell, EGT and others) they may affect the comm between blue box and USB isolator as well...

Meantime I may apply some shielding Technics used in aircraft piston engines.....

@pavlik-marinov. 

Yes, shield must only be connected in one end, that is very important, otherwise you can get electromagnetic induced current in the shield, which will cause problems. 

On all inputs and outputs there are components to protect the board. Inputs and outputs are not isolated. 

One thing that can be used is a USB isolator with separate power supply. This one for example: USB Isolator for Protection of all types of USB devices (usconverters.com)

Hello

Today I spent some hours troubleshooting and re-wiring. With some of my motors, the comm drops as soon the motor is started. with other the dyno works ok.

I noticed that 3 of 5 load cells has the cable shield connected to the cell body - bad
The one I have on the dyno as well. After cutting the shield, the comm get better.

Connecting the chasys to the water pipe helps as well.
It works in manual brake mode, but as soon I make a sweep, comm is lost.

Tried with New PVL coil and Denso IK31 6k ohm plug - the same - comm drops.

The most problematic is the EGT, standard sensor from BMW / Denso. It gas 300mm original sensr not shielded wire and then 1500mm shielded cable. no idea what to do here.

@Jostein. Do you have any noice suppressing components on the inputs, like aux, Lcell and K ???
Something need to be done here.
@Jostein - If you have some spare I / O on the MCU, can you try WiFi or BT conn to PC? This may solve the problems at once.
I volunteer for the tests.

Regards

P.

OK, tried to wrap aluminium folie around the sparkplugg wires....dont do that..sparks jumping between the wires and noise got worse ! After installing resistor plugs and isolate the wire from load cell about 3-400mm out from the cell the noise went down and the USB is more stable.

At max RPM USB signal on bluebox can flicker from green to yellow ...but not red ...at the end of the run but still not goes down.

So Im almost OK for now and will see whats happens in future.

This engine Im running now has distributor with mechanical points...mabe thats bad comp to eletronic pick up ?? What do you guys think??

Resistorized plugs are almost mandatory :-). I would go for that in any case! Cables are good to check too.

@yhakan  I always run the NGK plugs in all the VWs we see here.   You could try the resistor plugs but ill bet the cables are your issue.  I saw the same interference issues with the NGK plugs and the Taylor cables.  

@turbo-dub

Yes I use different ignition cables now, yellow siliconcables. But I realised that Im using "non resistor plugs" Bosch W 8 AC....mabe thats no god??

Should try NGK BR5HS instead! Should be simular to the Bosch but with resistor!

Thanks!

@yhakan  Like @admin stated the ignition is one major source of EMF.  If you haven't already checked this,  are you using different plug wires on the  engine since you last used your setup?  I have seen a number of problems with interference from plug wires that have improper EMF suppression (solid core wires, etc.) Especially on the AC VWs we see.  I had a 2276 I/C turbo that one big name brand of wires majorly interfered with the crank trigger/sensor signals to the ecu.    EMF from an ignition system can wreak havoc for sure.   MSD super conductor wire has worked well for me in our EFI set ups but you will have to custom make these for the stock socket style cap  or buy the tower adapters.  Just a thought. Have fun up there and  Happy New Year to you as well!

Thanks for all your help, Im going up to the mountains this coming week riding snowmobile...no electrical interference up there..but will continuing faultfinding after new year when back.

Curius how to protect Loadcell and wire from it to bluebox...Il read mail up there to...so Happy New Year to you all!

So I located the EMI disturbance the loadcell. When motor runs and the loadcell conneted USB dropps directly after start, so I routed the loadcell cable way out from all metal and connect to blue box...USB then stays on, but cable closer than ...approx 100-150mm to the metal, USB dies.

Another change since last time it all worked is that I replaced all rubber housings to COPPER tubes and pressure tube is running close to loadcell..se picture. Its def. related to spark noice! But I dont whant to reeroute the coppertube...so what can I do?? By just moving the laptop and all the electronics away ...more than 1 m ..didnt help.

Posted by: @harald-bergthaler

Is there a effect when disconnecting the AC adapter from the laptop ?
(so the laptop runs only by battery)

Tried that ..no effect.

It sounds like electro magnetic interference. 

Here are a few general tips on chasing electro magnetic interference problems:

• The most common source of disturbance on non-standard engines is the ignition. These can be extremely noisy. Dynomite dynos has a good guide, check it out: https://dynomitedyno.com/tech-corner/rfi-troubleshooting/ and follow the tips
• Variable frequency drives and power supplies can also be very noisy. Try for example with the stepper power supply off
• Ensure cables are routed away from noisy electronics (ignition and high current carrying cables) and as a second priority that they are as short as practically possible
• Move the instrument box away from noisy electronics
• You can ground the box, but only if there is no other ground in the system. You must avoid ground loops (i.e. two electrical paths to ground). A loop will generate currents from the electro magnetic radiation. Take care for external sensors, is the sensor's 0V is grounded? Then you already have 1 ground in the system, and must not add another
• Shields shall be open in the sensor end and connected to the YourDyno box shield input in the other end (otherwise, same problem as above). The load cell and RPM sensor are correctly made like this, but if you add your own sensors or modify the cable on the RPM sensor or Load cell, then ensure the shield is only connected to the YourDyno shield input

Let me know if you are stuck after checking the above, I can call you and we look at it together.

Is there a effect when disconnecting the AC adapter from the laptop ?
(so the laptop runs only by battery)

With aux 1 lamda and loadcell disconnected the USB is up and running with the motor running. Connect them to blue box while motor runs...Usb goes down ...no comunication!