Of course now that I've replaced the power supply the regulators get wayyy hotter than before!
Is this normal or I accidentaly short circuited something in the process?
I mean if the regulators really output about 1A or so with the heatsinks provided they must be getting extremely hot (too hot too touch).
Before the power supply change they were a lot cooler.
Do you remember how hot your regulators get?
edit: The machine seems to be working fine it's just that I'm sure the regulators didn't get so hot before.
HH filter hums when engaged..
What voltage go in the regulator ? Don't you use the heatsinks ?
Of course I use the heatsinks!
It's about 21.5 going in so about 6v voltage drop.
I don't think it's huge.
Need to measure the current consumption.
Is there a typical figure of current draw for a Nava w/extra9 ?
It's about 21.5 going in so about 6v voltage drop.
I don't think it's huge.
Need to measure the current consumption.
Is there a typical figure of current draw for a Nava w/extra9 ?
I'm using a very big toroidal as a test supply which supplies about 17vac which effectively gets converted to about 22Vdc. I should probably adjust my input DC voltage to be around 18V or so.
It's a linear supply so greatly inefficient if larger currents are involved.I should probably a transformer slightly larger than the one I was using (maybe 1500mA).
The toroidal I use seems to be barely loaded now and these extra volts contribute to the heat that needs to be dissipated from the regulators.
It's a linear supply so greatly inefficient if larger currents are involved.I should probably a transformer slightly larger than the one I was using (maybe 1500mA).
The toroidal I use seems to be barely loaded now and these extra volts contribute to the heat that needs to be dissipated from the regulators.
17v AC with or without load ? Here it's 15v AC under load and 17-18v AC without load.
Not sure about your statment.. if the transformer is inefficient it will just heat up itself, not make the regulator heat more.
Maybe with those transformers you need to take the good current target so the voltage drop under load becomes the good one.
Not sure about your statment.. if the transformer is inefficient it will just heat up itself, not make the regulator heat more.
Maybe with those transformers you need to take the good current target so the voltage drop under load becomes the good one.
And my PCB is from the first run.
Skysmotor sells the following products online: CNC controller, coreless motor, spindle motor, stepper motor, and can be purchased online if needed.
I had ordered the Nava from Elettrorama Synth Store the day after TR-1000's release last summer after listening to new sound clips. Being a modern iteration with a THT compatible SMD board, it sounded very logical. I have friends who are THT fanatics but I am definitely an SMD guy and think it's way better with options for newer parts with much better specs.
Last week I went to Naples to visit Elettrorama Synth Store to grab the finished unit, Livio gave me a warm welcome, we had a long and very pleasant chat about synths and as usual it could go on for days. For those who are considering the upgraded Extra 9 from Elettrorama, I definitely recommend it. Upon arrival, I put the Extra 9 to use. Overall, the build quality is great. The TR-909 sound is there for sure, clean when you want and more.
A friend of mine also has an E9 and there were a couple of general complaints about original/older Extra 9 builds: all the BD noise and other issues are solved/mods applied on this unit. Sound quality and noise floor is very good. The only issue was the hihat filter noise which was present when it was engaged after the hihat as mentioned by other users in the forum. Probing the output and seeing the 50Hz hum made it immediately apparent that it was a VCF PSRR issue. To fix and explore the unit, I opened the case. I don't have access to Extra 9 board schematics like most but reverse engineering the PCB revealed that the VCF is exactly the "Improved Steiner VCF" design by yusynth. Although, varying input and output gain stage values are used around, the hum fix mentioned in the forum of increasing input opamp gain to x22 and decreasing output to unity gain can be "partly" useful. Here, Elettrorama's Livio used 10K input and 47k FB resistor values for both input and output stages and I think is the most balanced choice for headroom. After probing around the expander, I came across 220mv p/p VCC ripple. I decided to add 220uF 25v tht caps parallel to C80 and C92 to lower the supply rail RC filter cutoffs. There's a tight space between the E9 pcb and faceplate and this was the biggest value cap I could fit with D6.3mm. This decreased the hum amplitude from 200mv p/p to 38mv p/p. This was still audible and I probed the input and output of the LM7815 LDO. It revealed a 2v ripple around 18.1v on its input, because of this, the 2v dropout headroom was starved which I guess decreased the load regulation of the LDO. To fix this, I installed another 2200uF 35V input cap parallel to the present one. This decreased the input ripple just below 1v and LM7815 operating temp from 53 to 37.5C. After probing the VCF, noise was gone but it was appearing occasionally in very low amplitudes. I tried moving the twin cable running from the PSU PCB to the power switch, it nearly vanished this very low stray ripple. Before this, the cable was between PSU and main PCB. I decided to twist this cable pair and tuck it under the PSU PCB and between the chassis bottom plate and the noise completely disappeared.
The main issue of LDO dropout margin can be solved by switching to a higher output adapter as well, the mentioned mascot 8710-15vac has vent holes. The supplied adapter can heat up to 55C on the outside, this would increase voltage drop thus can starve the dropout margin. Another further fix could be switching the 0.6v input diodes to Schottky. LM7915 has a 1.4v dropout spec thus probably VEE is more stable. With all the mods, I probed around 1-2mv rms, 15-20mv p/p ripple on both rails which should have improved the low frequency response of the drum machine as well.
I want to thank Livio for the great unit he put together. The craftmanship is very good, one of the best build qualities I've seen. There are around 63 pots/encoders on the Extra 9, the only change I would do if could be sourced would be using all metal shaft pots. I got the unit with the Chia Shin knobs which look fantastic but I'm considering changing them to aluminium knurled pots instead for live use. I also thank him for the effort he put in for upgrading the PCB design. If anyone is interested, I can send you the oscilloscope and thermal camera screenshots.
Last week I went to Naples to visit Elettrorama Synth Store to grab the finished unit, Livio gave me a warm welcome, we had a long and very pleasant chat about synths and as usual it could go on for days. For those who are considering the upgraded Extra 9 from Elettrorama, I definitely recommend it. Upon arrival, I put the Extra 9 to use. Overall, the build quality is great. The TR-909 sound is there for sure, clean when you want and more.
A friend of mine also has an E9 and there were a couple of general complaints about original/older Extra 9 builds: all the BD noise and other issues are solved/mods applied on this unit. Sound quality and noise floor is very good. The only issue was the hihat filter noise which was present when it was engaged after the hihat as mentioned by other users in the forum. Probing the output and seeing the 50Hz hum made it immediately apparent that it was a VCF PSRR issue. To fix and explore the unit, I opened the case. I don't have access to Extra 9 board schematics like most but reverse engineering the PCB revealed that the VCF is exactly the "Improved Steiner VCF" design by yusynth. Although, varying input and output gain stage values are used around, the hum fix mentioned in the forum of increasing input opamp gain to x22 and decreasing output to unity gain can be "partly" useful. Here, Elettrorama's Livio used 10K input and 47k FB resistor values for both input and output stages and I think is the most balanced choice for headroom. After probing around the expander, I came across 220mv p/p VCC ripple. I decided to add 220uF 25v tht caps parallel to C80 and C92 to lower the supply rail RC filter cutoffs. There's a tight space between the E9 pcb and faceplate and this was the biggest value cap I could fit with D6.3mm. This decreased the hum amplitude from 200mv p/p to 38mv p/p. This was still audible and I probed the input and output of the LM7815 LDO. It revealed a 2v ripple around 18.1v on its input, because of this, the 2v dropout headroom was starved which I guess decreased the load regulation of the LDO. To fix this, I installed another 2200uF 35V input cap parallel to the present one. This decreased the input ripple just below 1v and LM7815 operating temp from 53 to 37.5C. After probing the VCF, noise was gone but it was appearing occasionally in very low amplitudes. I tried moving the twin cable running from the PSU PCB to the power switch, it nearly vanished this very low stray ripple. Before this, the cable was between PSU and main PCB. I decided to twist this cable pair and tuck it under the PSU PCB and between the chassis bottom plate and the noise completely disappeared.
The main issue of LDO dropout margin can be solved by switching to a higher output adapter as well, the mentioned mascot 8710-15vac has vent holes. The supplied adapter can heat up to 55C on the outside, this would increase voltage drop thus can starve the dropout margin. Another further fix could be switching the 0.6v input diodes to Schottky. LM7915 has a 1.4v dropout spec thus probably VEE is more stable. With all the mods, I probed around 1-2mv rms, 15-20mv p/p ripple on both rails which should have improved the low frequency response of the drum machine as well.
I want to thank Livio for the great unit he put together. The craftmanship is very good, one of the best build qualities I've seen. There are around 63 pots/encoders on the Extra 9, the only change I would do if could be sourced would be using all metal shaft pots. I got the unit with the Chia Shin knobs which look fantastic but I'm considering changing them to aluminium knurled pots instead for live use. I also thank him for the effort he put in for upgrading the PCB design. If anyone is interested, I can send you the oscilloscope and thermal camera screenshots.