Say hello to my newest victim.
So far, I’ve replaced the stock power supply with a Jameco linear regulated power supply and the very crappy pot with a resistor ladder attenuator.
Currently, I’ve removed the relay closest to the pot, to make room for the new pot, but to avoid burning down my house can anyone confirm I’m bypassing it correctly?
Below is the pinout of the relay:
Using the above I'm guessing I just need to short these 2 pins on either side
The thing is the 789 has relay's up the wazoo. I figure this is the reason I can't check if I'm connecting the right pins, because the signal from the RCA jacks to this relay isn't actually connected, from another relay I'm assuming, until this thing powers on.
Am I doing this right? I've desoldered the relay and am just waiting for the go ahead to short pin 2 to pin 4 and pin 9 to pin 7. I know for sure that pin 2 and pin 9 are shorted to the attenuator's input for their respective channel.
Lastly, near the 24V input voltage there is a rail splitter to two 12v rails and the original pcb has space for two smoothing capacitors. I’m willing to bet they were removed due to cost constraints. Would there be any benefit to adding these back in, and what value capacitors I should put in?
Thanks!
So far, I’ve replaced the stock power supply with a Jameco linear regulated power supply and the very crappy pot with a resistor ladder attenuator.
Currently, I’ve removed the relay closest to the pot, to make room for the new pot, but to avoid burning down my house can anyone confirm I’m bypassing it correctly?
Below is the pinout of the relay:
Using the above I'm guessing I just need to short these 2 pins on either side
The thing is the 789 has relay's up the wazoo. I figure this is the reason I can't check if I'm connecting the right pins, because the signal from the RCA jacks to this relay isn't actually connected, from another relay I'm assuming, until this thing powers on.
Am I doing this right? I've desoldered the relay and am just waiting for the go ahead to short pin 2 to pin 4 and pin 9 to pin 7. I know for sure that pin 2 and pin 9 are shorted to the attenuator's input for their respective channel.
Lastly, near the 24V input voltage there is a rail splitter to two 12v rails and the original pcb has space for two smoothing capacitors. I’m willing to bet they were removed due to cost constraints. Would there be any benefit to adding these back in, and what value capacitors I should put in?
Thanks!
Yah, sorry scratch all of the above about the relay. With some probing with the amp on I found none of the pins connect to the RCA inputs. If some of the pins are shorted via a multimeter in ohm mode to ground the amp does briefly go into protection mode. I think given the proximity to the source selection button this relay just mutes the output when you switch between RCA and balanced inputs. The amp works fine with it completely removed. Scratch all my assumptions and idiocy above.
So, about those missing caps, any ideas?
So, about those missing caps, any ideas?
Hi Tom, this is indeed the original run Massdrop THX 789. The place for missing caps is at the top of the PCB near the rail splitter. In the photo it's between the balanced inputs and the 24V power supply input. When I probed the via's for the missing caps I saw 12 Volts across each cap and they are not in series or parallel with each other. They are completely separate.
Each cap appears to be in parallel with a 12V input voltage. This leads me to believe they are smoothing caps and I would need at least 15V caps. I'm not sure how much capacitance is needed. I was warned on SBAF by a friendly member that too much capacitance could cause ripple. I was also warned on ASR that it could create too much in rush current, so I just want to be wary of both of these. Thank you for any input you can provide!
Each cap appears to be in parallel with a 12V input voltage. This leads me to believe they are smoothing caps and I would need at least 15V caps. I'm not sure how much capacitance is needed. I was warned on SBAF by a friendly member that too much capacitance could cause ripple. I was also warned on ASR that it could create too much in rush current, so I just want to be wary of both of these. Thank you for any input you can provide!
So, I'm going to try to approximate the capacitance using the following method and 2 assumptions.
1) The missing cap is a Rubycon as the rest of the electrolytic are.
2) The voltage is 25V.
I plan to use the diameter of the silkscreen on the PCB to approximate the size of the capacitor that Rubycon makes that would fit. Is this approach really stupid?
1) The missing cap is a Rubycon as the rest of the electrolytic are.
2) The voltage is 25V.
I plan to use the diameter of the silkscreen on the PCB to approximate the size of the capacitor that Rubycon makes that would fit. Is this approach really stupid?
16V caps would make sense for a 12V supply. Electrolytic cap voltages tend to follow the R5 range of values, 10/16/25/40/63
If they made 12V caps they would do too as electrolytic cap voltage ratings are working voltages, not absolute maximum.
16V cap of the same volume as a 25V cap would be about 2.5 times the capacitance for the same range of capacitors, volume scales with energy storage which scales with C * V^2
If they made 12V caps they would do too as electrolytic cap voltage ratings are working voltages, not absolute maximum.
16V cap of the same volume as a 25V cap would be about 2.5 times the capacitance for the same range of capacitors, volume scales with energy storage which scales with C * V^2
No, other way round, more capacitance = less ripple. The capacitive impedance and ESR will both drop with higher values of capacitance.
I confirmed this morning that the missing caps are indeed in parallel with the existing 3300uF 16V caps. Would it be safe to add additional 6800 uF 16V caps in parallel, or do I have to be concerned about too much additional capacitance causing oscillation? Would in rush current be an issue? TIA!
Please pardon my ignorance, I really appreciate your help. I believe the caps I plan to install can withstand a ripple current of 1.9A at 16V, 6800uF, in parallel with caps that can withstand 1.2A at 16V, 3300 uF. My power supply is a linear regulated one that can only provide 12 watts, so does that mean I'm good to go? I mean, there's no way the power supply could provide more than what my caps and withstand right?
After some reading this morning, I also intend to add a 0.1uF bypass capacitor to a 6800uf 16V cap. My thinking is as follows:
1) The 789 is known and measured to have noise coming from the power supply mess with the circuit when left ungrounded. The only ground it gets is when balanced inputs are used. Single ended only can mean a noisy supply.
2) The 789 is notoriously known for fatiguing sound. To me this usually means high frequency noise.
3) Switching to a lps improved the micro dynamics of the amp. This lps has much lower v-ripple than the original switcher.
4) The quality of the psu is hampered by the rail splitter it needs to go through for plus/minus 12V which can cause its own noise/distortion.
5) Other thx feed forward amps like the hpa4 don’t suffer from this.
I suspect by tripling the capacitance I can reduce the ripple by over a factor of 2 and adding a bypass cap will allow for even more smoothing at high frequency’s. My hope is with cleaner 12 volt rails I can improve the microdynamics and macrodynamics.
Feel free to chip in if I’m being retarded. I won’t be offended.
1) The 789 is known and measured to have noise coming from the power supply mess with the circuit when left ungrounded. The only ground it gets is when balanced inputs are used. Single ended only can mean a noisy supply.
2) The 789 is notoriously known for fatiguing sound. To me this usually means high frequency noise.
3) Switching to a lps improved the micro dynamics of the amp. This lps has much lower v-ripple than the original switcher.
4) The quality of the psu is hampered by the rail splitter it needs to go through for plus/minus 12V which can cause its own noise/distortion.
5) Other thx feed forward amps like the hpa4 don’t suffer from this.
I suspect by tripling the capacitance I can reduce the ripple by over a factor of 2 and adding a bypass cap will allow for even more smoothing at high frequency’s. My hope is with cleaner 12 volt rails I can improve the microdynamics and macrodynamics.
Feel free to chip in if I’m being retarded. I won’t be offended.
Since the functionality of the relay isn't really known, wouldn't it be better to simply move it by adding wire leads and setting it out of the way of the pot upgrade?
The relay is there for a reason, and that was enough to justify its cost to the manufacturer. The amp may seemingly "work" fine without it, but there is no doubt some functionality will be lost.
I know this type of product are challenging to mod since schematics aren't usually available so good luck.
The relay is there for a reason, and that was enough to justify its cost to the manufacturer. The amp may seemingly "work" fine without it, but there is no doubt some functionality will be lost.
I know this type of product are challenging to mod since schematics aren't usually available so good luck.
My thoughts were exactly yours until I found that the amp no longer mutes when switching from balanced to singled ended inputs. I found this out completely by accident when I was probing the amp while on with my multimeter and accidentally put it into protection mode for a second while it was in ohm mode checkout out the pads I left bare.