Thandar TG501 earth bonding

[tron9000]

I'm in the process of testing a Thandar TG501 signal generator. Part of my testing is to do a PAT inspection on it as i'm selling it on. Need to make sure its safe.

I performed an earth bond test on it with a test current of about 10A and it failed. The way it failed though was telling: it basically had an earth bond resistance of 0.1R, then went up and up till it went practically open circuit! Indicating that the earth bond couldn't hold the 10A current and and clearly not liked the ordeal.

Sure enough when I cracked it open, there was a lifted charred track. Tracing this track, it when to the outer of the 50R output BNC on one side of the break and to chassis on the other. The BNC on closer inspection is isolated from the metalwork withe a plastic collar.

Now I've PAT inspected signal generators before by testing at the BNC shield, and they have survived the ordeal, but this is a new one on me cos the BNC outer goes to earth via 0V which is connected to the mains earth (So says schematics in the service manual). Yet there is an earth wire directly to the chassis. And all BNC's are like this on it.

Question is: do I repair the break and leave it as it was designed or do I solder a wire from mains earth to all BNC Outers?

[Chris56000]

Hi!

Thandar TTi gear should NOT be earth–bond tested between the mains earth and any of the signal earths – for the reason you unfortunately discovered when the track burnt–out!

Repair the burnt–track then test ONLY between the rear case metalwork and the end of the mains earth lead, etc., then do an ordinary ohms continuity test using a DMM between BNC 0V and mains earth – this is all that is necessary for the Generator to be safe!

Chris Willams

[bd139]

To note there is usually a strap between the IEC socket boot and the front BNCs on the later TG200 series. The 501 is an older unit from Thandar’s shonky as feck era.

Tek recalled the TDS200 series over this precise flaw as the rest of the instrument is floating so if that trace burns out then it can result on mains potential on the BNCs.

[tron9000]

That was my concern: if the earth bond breaks like it did, then BNC outer is floating and should be grounded, both for functionality and for safety.

Least now I know its a known issue. cheers

[Chris56000]

Hi!

As an alternative, a 2M2 high–voltage VR37 type resistor connected between BNC outers/Signal 0V and the incoming mains–earth terminal will allow the signal earths safe isolation from the mains whilst allowing static voltages or stray voltages from SMPSUs in associated equipment the unit is connected to to safely discharge back to mains PE!

Chris Williams

[tron9000]

I mean that would be great for those application, and ESD and voltage isolation.

But in the event the BNC outer becomes live and shorts to mains PE, its relying on the PCB traces to carry that current before the MCB trips. It won't be there the second time it happens. 2M2 would allow a conduction path but nowhere near the limits required from a class 1 earth bond test.

Or have I miss understood?

[IanMacdonald]

You should not do high current earth tests to any SIGNAL socket on test equipment, or on computers, audio gear etc. Signal connectors are not meant to be used as earth bonding points. The outer of a coax is intended as an shield for the signal line. It may or may not be connected to mains earth depending on the equipment design. Either way, it is not meant as a facility for earthing other equipment to.