Substitude CH340G with CH340C

Post Reply
dc8lz
 

Posts: 39
Joined: 31 Dec 2021, 01:35

Substitude CH340G with CH340C

Post by dc8lz »

Hello!

Sometimes CH340G is out of stock at jlcpcb. What about replacing it with a CH340C... Chrystal Y3 may be useless then... but possible? Or do we brake something?

73, Stephan
mm0gyx
 

Posts: 83
Joined: 30 Dec 2021, 20:22

Re: Substitude CH340G with CH340C

Post by mm0gyx »

dc8lz wrote: 14 Jan 2022, 14:33 Hello!

Sometimes CH340G is out of stock at jlcpcb. What about replacing it with a CH340C... Chrystal Y3 may be useless then... but possible? Or do we brake something?

73, Stephan
I had to get them from LCSC and solder them separately, they claimed to have lots, but then I got an email saying we have only 35. Don't get any from aliexpress, I did it and several of them are useless, actually I have one to desolder with the heat gun, not looking forward to it. The others I just pressed on the board with fingers to do a quick test.

73,

Ian MM0GYX
dc8lz
 

Posts: 39
Joined: 31 Dec 2021, 01:35

Re: Substitude CH340G with CH340C

Post by dc8lz »

...so you got CD340C or just speak about CH340G?
mm0gyx
 

Posts: 83
Joined: 30 Dec 2021, 20:22

Re: Substitude CH340G with CH340C

Post by mm0gyx »

dc8lz wrote: 14 Jan 2022, 15:07 ...so you got CD340C or just speak about CH340G?
Sorry, I caused confusion, I'm talking 340G, original.

I just desoldered my bad one, microsurgery.
dc8lz
 

Posts: 39
Joined: 31 Dec 2021, 01:35

Re: Substitude CH340G with CH340C

Post by dc8lz »

https://www.ebay.co.uk/itm/154363637553

good enough for desoldering jobs like this ;)
mm0gyx
 

Posts: 83
Joined: 30 Dec 2021, 20:22

Re: Substitude CH340G with CH340C

Post by mm0gyx »

dc8lz wrote: 14 Jan 2022, 15:40 https://www.ebay.co.uk/itm/154363637553

good enough for desoldering jobs like this ;)
Yes, I have a friend who has one, I was going to borrow it, but I used a needle to heat the pad and carefully lift the leg. I repeated it 15 times and it was off. Yes, I know it is risky, but the footprint in this case was unharmed, I cleaned up wick solder wick and fitted the new one.

I was worried the heat gun might also desolder some of those tiny 0603 parts nearby.

This time I got lucky.

73,

Ian MM0GYX
DL2MAN
 

Posts: 377
Joined: 30 Dec 2021, 19:18
Contact:

Re: Substitude CH340G with CH340C

Post by DL2MAN »

I seriously cannot tell if CH340G can be substituted with CH340C. We´ve never tested this.
The G Version (with crystal) was chosen, assuming that a crystal would create less Noise compared to an undefined OnChip RC Oscillator.
I have nothing to back this up, and it was just a safety decision.

You could try it and share your results.

73 Manuel; DL2MAN
mm0gyx
 

Posts: 83
Joined: 30 Dec 2021, 20:22

Re: Substitude CH340G with CH340C

Post by mm0gyx »

DL2MAN wrote: 15 Jan 2022, 10:02 I seriously cannot tell if CH340G can be substituted with CH340C. We´ve never tested this.
The G Version (with crystal) was chosen, assuming that a crystal would create less Noise compared to an undefined OnChip RC Oscillator.
I have nothing to back this up, and it was just a safety decision.

You could try it and share your results.

73 Manuel; DL2MAN
I was thinking of trying it, but I'm going away with work for two weeks so it needs to wait, then I need to try to find the 'c' version somewhere locally.

Legs 7 and 8 need to be lifted off the board for a quick test. If I do it (before anyone) I'll share results.

73,

Ian MM0GYX
mm0gyx
 

Posts: 83
Joined: 30 Dec 2021, 20:22

Re: Substitude CH340G with CH340C

Post by mm0gyx »

mm0gyx wrote: 16 Jan 2022, 12:48
DL2MAN wrote: 15 Jan 2022, 10:02 I seriously cannot tell if CH340G can be substituted with CH340C. We´ve never tested this.
The G Version (with crystal) was chosen, assuming that a crystal would create less Noise compared to an undefined OnChip RC Oscillator.
I have nothing to back this up, and it was just a safety decision.

You could try it and share your results.

73 Manuel; DL2MAN
I was thinking of trying it, but I'm going away with work for two weeks so it needs to wait, then I need to try to find the 'c' version somewhere locally.

Legs 7 and 8 need to be lifted off the board for a quick test. If I do it (before anyone) I'll share results.

73,

Ian MM0GYX
Right, I did it. First the results:

A pre-mod (340G) mainboard with LM833 op amp has the following antenna disconnected noise floor: 80m -116 dBm, 60m -117 dBm, 40m -115 dBm, 30m -106 dBm, 20m -108 dBm

A post mod (340C) mainboard, same op amp and using the same RF board antenna disconnected is:
80m -107 dBm, 60m -106 dBm, 40m -107 dBm, 30m -105 dBm, 20m -106 dBm

So, on the face of it, at lower frequency the G looks better than C. Both radios have various internal noises. I tested with an S9/S2 signal generator, in both cases the S2 can clearly be heard. I'm not sure of the difference if I had installed a lower noise op amp. For my part, it will do.

In terms of fitting the device, I snapped off the legs 7 & 8 at the edge of the IC. Then soldered it in and powered up, I got a USB device not recognised warning, so I removed the driver and ensured it was definitely from WCH, same problem. A quick google suggested I replace the 1 nF capacitor C34 with a 100 nF version, that actually worked. It is attached to pin 4 (V3), the datasheet calls for 100 nF.

Anyway, I guess I'm happy with that result. In my case I ordered the PCBs without CH340G, a mistake, they came back in stock a couple of weeks later. I ordered some from LCSC, but they didn't have enough, hence the "C" version also.

73,

Ian MM0GYX
DL2MAN
 

Posts: 377
Joined: 30 Dec 2021, 19:18
Contact:

Re: Substitude CH340G with CH340C

Post by DL2MAN »

Hello Ian,

thanks for this experiment.
Even though it will probably still work in practice because of high noise floor on our short wave bands anyway (more noise on lower bands anyway), I find it not really desireable to give up 2-7dB Sensitivity.
But it might serve as expedient solution if you have no other choice.

73 Manuel; DL2MAN
mm0gyx
 

Posts: 83
Joined: 30 Dec 2021, 20:22

Re: Substitude CH340G with CH340C

Post by mm0gyx »

DL2MAN wrote: 01 Feb 2022, 11:45 Hello Ian,

thanks for this experiment.
Even though it will probably still work in practice because of high noise floor on our short wave bands anyway (more noise on lower bands anyway), I find it not really desireable to give up 2-7dB Sensitivity.
But it might serve as expedient solution if you have no other choice.

73 Manuel; DL2MAN
No problem, I did it because I had to. I'm not convinced the loss of the sensitivity is down to the CH340, but I have no other ideas, so it's just my hunch.

73, Ian MM0GYX
mm0gyx
 

Posts: 83
Joined: 30 Dec 2021, 20:22

Re: Substitude CH340G with CH340C

Post by mm0gyx »

mm0gyx wrote: 01 Feb 2022, 13:56
DL2MAN wrote: 01 Feb 2022, 11:45 Hello Ian,

thanks for this experiment.
Even though it will probably still work in practice because of high noise floor on our short wave bands anyway (more noise on lower bands anyway), I find it not really desireable to give up 2-7dB Sensitivity.
But it might serve as expedient solution if you have no other choice.

73 Manuel; DL2MAN
No problem, I did it because I had to. I'm not convinced the loss of the sensitivity is down to the CH340, but I have no other ideas, so it's just my hunch.

73, Ian MM0GYX
Oh, I forgot to say, I did find a huge birdy at 7060 khz, about S8, but this was there with 340C and 340G rigs.
mm0gyx
 

Posts: 83
Joined: 30 Dec 2021, 20:22

Re: Substitude CH340G with CH340C

Post by mm0gyx »

DL2MAN wrote: 01 Feb 2022, 11:45 Hello Ian,

thanks for this experiment.
Even though it will probably still work in practice because of high noise floor on our short wave bands anyway (more noise on lower bands anyway), I find it not really desireable to give up 2-7dB Sensitivity.
But it might serve as expedient solution if you have no other choice.

73 Manuel; DL2MAN
Hi Manuel,

Further to my little test I realised I wasn't comparing the same rig, pre and post mod, so not good experimental process, how do I know source of rx floor difference is due to 340? I don't.

Today I took more care as follows:

My own (tr)uSDX powered from 13.8 volt shack supply and also plugged in to PC in both cases.
SMA terminated with 50 ohm SMA load.
My rig uses LM833 op amp, I am planning to replace it with less noise version at some stage to compare results.
Headphones plugged in.

First test using my rig with 340G:

80m -116 dBm
60m -118 dBm
40m -117 dBm
30m -107 dBm
20m -114 dBm

Then I used the rework station to remove C42, C43, Y3 and U5. I also removed C34 and replaced it with 100 nF type. U5 was replaced with 340C device. Then exact same unit, mine, was again powered up with same parameters and exact same frequencies, results:

80m -117 dBm
60m -118 dBm
40m -118 dBm
30m -108 dBm
20m -114 dBm

I tuned through the whole band and didn't find any QRM with the C version that wasn't already there with the G version. Actually there's some big birdies around, but always in both cases.

I'm not trying to convince you to change G to C, I just wanted my experimental method to be cleaned up and give a more accurate set of results. I hope it might be useful to anyone.

Cheers, 73,

Ian MM0GYX
Post Reply