Genesis 1 & Mega Drive 1 stereo mod no headphones volume control

[Firebrandx]

Excellent! Does it also properly mix in the PSG as well doing that? I assume it does, but wanted to ask.

Yep! PSG is fully intact. This means all I need to do is get the proper input caps and resistors going and this amp will blow away the CXA1034! I've asked both Ste and Ace for their input on what to use for the input side of the amp now that the DC offset is measured at 0.01V for both L and R channels. I'm additionally going to install a socket in place of the CXA1034 so I can simply plug-and-play it whenever I need to test the chip for whatever reason.

I can also hardly wait to try this out with the cap replacement kit I assembled. It's on its way from digikey and should arrive today if I'm lucky and the post office doesn't put it off for tomorrow's delivery. I will immediately recap the entire system as soon as it gets here.

[CrossBow]

You know...if you're going to socket in the 1034 for testing and comparison, couldn't you just wire in a switch to enable or disable the +5v to the 1034 as needed so you don't have to wear out the socket or the pins on the chip pulling it and inserting it so much?

[Firebrandx]

You know...if you're going to socket in the 1034 for testing and comparison, couldn't you just wire in a switch to enable or disable the +5v to the 1034 as needed so you don't have to wear out the socket or the pins on the chip pulling it and inserting it so much?

Possibly, but I don't expect to be plugging in the chip all that often (if at all). If everything works as planned, I won't even need the CXA1034 any more. The mod board does a much better job of amplifying the audio without all the messy distortion. I just wanted to put a socket in there just in case down the road I want to put the CXA back into the machine.

[Firebrandx]

Update:

In order to increase the volume on the board, I just need to raise the resistor value for R5 and R6. Just tried with 400k and it's almost back to the same volume level (will keep adjusting until I have it perfect). Edit: 604K Ohms nailed the sweet spot, and I've already sourced a Panasonic resistor for this part.

Additionally, I compared the "Green Hill Zone" track with the VGM version and they sound almost identical! The only thing I need to change is the high frequency filtering (the cutoff isn't low enough compared to the VGM version), and possibly add a hair more bass response. Other than that, this is getting close to perfection! But the trick most definitely is removing the CXA when using this mod board. The CXA introduces all that 'fuzzy' noise and messes up the frequency cutoffs, and removing it has made a world of difference.

[Firebrandx]

Okay here's a recording with known PSG audio in it (Sonic Green Hill Zone) taken from my amp board with the CXA1034 removed from the chain:

https://filetrip.net/dl?XMb0u85lin

I need some opinions on the quality. Should I add in more bass? Less high frequency? Both? Or leave it as is? Keep in mind this is coming from my VA3 where I had previously raised the cutoff for high frequency by replacing a couple of the SMD caps on the Genesis itself.

[CkRtech]

Looks like you've got some clipping in your waveform. Load it up in audacity and check out the high side. Easier to see when the countermelody kicks in at about the 16 second mark.

I experienced the same issue when working on the "TL072-altered" circuit and had to adjust the VRef value as it was too low.

[Firebrandx]

Looks like you've got some clipping in your waveform. Load it up in audacity and check out the high side. Easier to see when the countermelody kicks in at about the 16 second mark.

I experienced the same issue when working on the "TL072-altered" circuit and had to adjust the VRef value as it was too low.

I'm not seeing what you're talking about. The flac file isn't showing any clipping.

Here's just before 16 seconds and to 18 seconds. Nothing is getting clipped:

[Xaranar]

I can hear some sort of hiss in the background, though that might just be the DAC. I do think it sounds a little flat and could use some more bass response.

[Firebrandx]

I seem to be also losing stereo separation. It's still there to some degree, but not as distinct as before. I may have to scratch the new approach and go back to using the CXA and just try to adjust the bandpass for highs and lows. The new cap kit arrives today, so I'll reinstall the CXA long with all the new caps and start over.

[CrossBow]

I'm not seeing what you're talking about. The flac file isn't showing any clipping.

Here's just before 16 seconds and to 18 seconds. Nothing is getting clipped:

Is it just me...or is that waveform inverted? I will need to record Greenhill Zone off my va3 and see. I know most people will say it doesn't matter, but to my ears I can hear it as it is similar to the phase being wrong when you have your wiring on your speakers backwards. Will have to listen to this more with my HD600s when I get home after work.

[Firebrandx]

It is inverted, but it doesn't matter. I've scrapped that experiment.

However, before the complete recapping I'll be doing later today, I'd like you guys to try this blind audio test from Streets of Rage:

https://filetrip.net/dl?eaNiNyoOm4

Tell me how "A" was captured and how "B" was captured. Also please describe any difference you perceive in quality and equalization (not with equipment, but rather with your ears).

[CkRtech]

I'm not seeing what you're talking about. The flac file isn't showing any clipping.

See how the top of the waveform doesn't seem to get past +0.4 whereas the bottom hits -0.7?

[CrossBow]

It is inverted, but it doesn't matter. I've scrapped that experiment.

However, before the complete recapping I'll be doing later today, I'd like you guys to try this blind audio test from Streets of Rage:

https://filetrip.net/dl?eaNiNyoOm4

Tell me how "A" was captured and how "B" was captured. Also please describe any difference you perceive in quality and equalization (not with equipment, but rather with your ears).

I honestly can't tell how you captured them. But "B" sounds better to me. The bass is still strong without being over bearing as it is with "A". And it seems like the stereo separation is better with "B" as well. I also was able to hear more hiss in the drum samples from "A" vs "B".

So listening with my sound setup here, I thought "B" sounded better and more clear overall.

Does that help?

[Firebrandx]

See how the top of the waveform doesn't seem to get past +0.4 whereas the bottom hits -0.7?

That has nothing to do with clipping. Clipping is when your peak volume goes past 1.0 or -1.0, and the envelope literally chops it off. Now aside from the wave form being inverted, what you're talking about is DC offset. If the DC offset isn't perfectly calculated, the peaks of the wave will favor either either negative or positive phase. That's something that can be corrected on the board, but I'm still working on how I want to approach the mod. It may ultimately end up being that the best mod is to simply tap the headphone output lines and route those to rear RCA jacks, in which case, the point of the mod is to merely provide a more stable connecting with RCA jacks over the mini headphone jack (which is prone to getting dirty and causing static dropouts). If that's the case, then the mod board itself was a waste of time and money on my part, but it's worth finding these things out.

This past evening I spent a few hours doing the full recap with the best Nichicons I could find. I also installed the dual wipe socket for the CXA1034, and installed two brand new voltage regulators. We'll see if this has any effect on the output tests tomorrow. For now, here's a pic of my work:

[CkRtech]

That has nothing to do with clipping. Clipping is when your peak volume goes past 1.0 or -1.0, and the envelope literally chops it off. Now aside from the wave form being inverted, what you're talking about is DC offset.

Clipping can happen regardless of where it is on the scale. It doesn't have to happen at the extremes like that. There can also be hard clipping as well as soft clipping. The DC offset of the recording provided in post #185 is fine, but the waveform contains partial clipping.

It is similar problem that I had and can be found starting here - http://www.sega-16.com/forum/showthr...l=1#post774794 Ace's post below it mentions the waveform being partially clipped. That is what I meant in my response yesterday.

And one of the "in progress" images in one of the later posts shows a screenshot comparison of the difference in the waveform when I adjusted the Vref value. You can see the different voltages as well as the change in the waveform - how it became more symmetrical here - http://www.sega-16.com/forum/showthr...l=1#post775086

But - positive vibes FBX! I admire your dedication.