search

sinara-hw / Pounder

PDH/phase lock signal generator for Stabilizer
5 stars 4 forks source link

v1.1 changelog #86

Closed gkasprow closed 10 months ago

gkasprow commented 3 years ago
jordens commented 3 years ago

Great. Could you spin the RC production files so we can give it a quick check?

gkasprow commented 3 years ago

done

jordens commented 3 years ago

I had a look at the changes. Apart from the opamp change #79 I'm ok with v1.2 and would like to proceed with getting it manufactured. As a note I still think that:

But since this is all still experimental, I'm not going to fight these.

jordens commented 3 years ago

@gkasprow AFAICT there is only the opamp change (#79) pending for v1.2. Can we cut the release and send the production files?

gkasprow commented 3 years ago

The baluns should all become TC1-1T (replace 7 ADT1-1WT)

Just to simplify the BOM?

The use of pricey "special RF" capacitors for DC-blocks at 1-200 MHz is not supported by data, evidence, or experience with the other dozen designs that don't have and don't seem to require them. When I look at standard MLCC for DC-blocking in 1-200 MHz, they are all just fine.

I was curious about real benefits. We can always use Hi-Q caps instead.

Putting RF output and RF input signal paths into the same RF shield box makes the purpose and benefit of the boxes unclear to me: is it for emission or immission or crosstalk?

Only output path attenuator was placed under the shield, the PA sits on the other PCB side. If we discover crosstalk, I will place the attenuator on the bottom PCB side

The RF output amplifiers can go away.

Why?

jordens commented 3 years ago

Just to simplify the BOM?

BOM simplification, availability (the ADT1-1WT are quite a bit more exotic and hard to replace with anything footprint compatible, while the TC1-1T style is ubiquitous), and board space.

If we discover crosstalk

Right. We should measure crosstalk.

The RF output amplifiers can go away.

They are typically not needed. Like for Urukul and Mirny, the usual amplifiers (e.g. Booster) would take 0 dBm (-5 dBm might be a little low however). And power usage: getting rid of them would reduce the power consumption by 0.7 W.

gkasprow commented 3 years ago

The amplifier can be bypassed by the assembly option. I will make the bypass default configuration.

dtcallcock commented 3 years ago

They are typically not needed.

Actually I find the -5 to 10dBm range is often what you need to add few percent sidebands to your laser with a modulator device/laser input, and having to add an external amp to achieve these very modest signal levels would be annoying. But as long as it's retained as an option then that's fine.

Noting on the schematic/wiki that an ERA 50SM+ can be populated here instead would also be nice. This extends the output to an even more useful +13dBm. Sure it's another 280mW of power consumption but it's just a suggestion so nobody needs to compromise.

gkasprow commented 3 years ago

v1.1 release is ready

gkasprow commented 3 years ago

@kaolpr @akaminska

jordens commented 3 years ago

Thanks!

dtcallcock commented 3 years ago

image Super minor, but can you make R34 the omitted component, not L15? Any reasonable lab will have a kit with suitable resistors but the right inductor would probably have to be ordered.

gkasprow commented 3 years ago

the inductor will make a stub

dtcallcock commented 3 years ago

true dat

gkasprow commented 3 years ago

One would have to order the amplifier anyway :)

dtcallcock commented 3 years ago

Good point. Perhaps add the part number as a schematic annotation so one can order it without having to find an Altium licence though?

dnadlinger commented 3 years ago

Shouldn't we be able to just flip them around? A 82 Ohm resistor should be low-enough Q at our frequencies? Just asking out of curiosity.

gkasprow commented 3 years ago

yes, we can swap them. The resistor has smaller pads than the inductor so capacitance should be not an issue obraz