Re: Tek cam switch contact current rating


Harvey White
 

That was an interesting experiment just done.


I took a resistance decade box, a 7A26, and just held a wire to the ring on the channel 2 input with the gain set at 1 volt/div.


0-500 ohms reads identify

600 to 3200 reads x10 (see note)

3300 to 4700 reads X1

4800 to 7800 reads X100

7900 to 16000 reads X10

17000+ reads X1

I did not find a gain setting of X 0.1

However, changing the range (note) changes the displayed value depending.

600 to 3200 looks like it ought to read X1000, but is modified by the range factor to a more reasonable value

Ditto for 3300 to 4700.

If I put together a suitable fixture, I'll do more experiments.

On the switch rating, no idea.  I rather suspect that 50 ma is more likely, and if it does 300 ma, it's only a carry, not a make/break.

Harvey

On 10/9/2020 3:08 PM, Ed Breya via groups.io wrote:
Does anyone happen to know the actual contact current rating of the Tek cam switches commonly used in many front-end attenuators and other spots? I'm modifying a 7A26 to be used specifically for P6021 current probes, and some big high current, low frequency clamp-on probes. The unit will no longer be a conventional plug-in, but will have low impedance input terminations, frequency compensation, and amplifiers, and fixed for current indication in A and mA, scaled in the same 1-2-5 sequence, matching the front panel numbers and on-screen readout. It will be kind of like the 7A14, but with two channels. In the 7A14, it appears the same kind of cam switch contacts are used in the low impedance circuits up to maybe 10 mA or so, but I'd like to know how high they can go, because it gives me more options on re-tasking the contacts for various purposes.

In order to avoid re-coding or redesigning all the readout circuits, the present plan is to keep the attenuator assemblies mostly intact, leaving the X2 and X5 1 meg attenuator tub stages in place, and isolating and removing the two X10 stages from the signal path, so the contacts can be decoded for driving external relays for all the low impedance routing and scaling. Another option is to use the X10 contacts directly on the signals, with tubs modified for low impedance - around 10 ohms in this case. This could get very tricky and have problems with isolation, so most likely it will be all done with relays.

The end result will be that the modified and added circuitry will handle the current signal processing, scaled in proper decades, and feed the remaining 1 meg input section for the final 1-2-5 sequencing. Since this is all AC-coupled, the AC-GND-DC switch for each channel isn't needed, and will be for other uses like probe type selection. These switches also use the same type contacts, that could be used directly for passing low impedance signals, or generating control signals.

So anyway, I'd like to find out if there are any current specs on these contacts. The largest internal process signal currents I can picture may be in the tens of mA. The largest possible applied current will be from the big probes, which may reach over 300 mA peak, if hundreds of amps pass through the clamp. No matter what, there will be protection built into the front ends so the inner circuitry never sees this much. This level will only be handled by relays.

Overall, it looks pretty straightforward. The basic ranges will be 5A-5mA per division with the P6021s, and 50A-50mA with the big probes. One very convenient thing about keeping the original readout intact, is the automatic X10 scaling used for 10X voltage probes, via the code ring. This will instead be actuated internally when the big probe type is selected. Actually, this leads to another question. Does anyone know if the readout can automatically scale the other way, showing 10X more sensitivity? Presuming this all works out OK, I'll probably want to study options for ten times more gain at the low end, to reach say a 500mA-500uA range with the P6021. It would be nice if the readout could easily shift down as well as up. I know it does X10 and X100 (and maybe X1000), but how about 1/10 and such?

Ed




Join TekScopes@groups.io to automatically receive all group messages.