testing epic controller on antminer s19xp

I'm running the power setting at 120th. The meter showed 970.37 kW at 12:06 am. I’ll check back in the morning with wattage updates. At 1:06 am, it was 972.81 kW. So that’s 2.44 kW for 120th. That breaks down to about 20.33 watts per TH. I’ll grab more readings in the morning. At 9:06, it’s at 993.17 kW. In 9 hours, that's 22.80 kW total. Doing the math: 22.80 / 9 = 2.5333 kW for 120th. When I divide 2533 by 120, I get around 21.11 watts per TH. Not much savings compared to the rated 21 watts per TH. But I have complete control over the fans now. They’re set to 61%, about 4900 RPM, so they’re super quiet in my garage. This setup lets me run the S19XP without a ton of noise. Update: At 6:06 pm, it read 1016.59 kW. Earlier it was at 970.37 kW at midnight. That means 46.22 kW over 18 hours, which equals 2567 watts per hour. 2567 / 120 gives 21.39 watts per TH. It’s interesting how the high-end meter rates it at 21.39 watts per TH while the software says 23.8 watts per TH. So I’ve been running at 120th for 18 hours and really haven’t improved efficiency much, but the fan speed is down to 61% and the garage noise is manageable. Now I’ve tuned it down to 70th, with the lowest setting being 60th. So far at 8:00 pm, the kW reading is 1019.30. By 10:00 pm, it hits 1022.31, totaling 3010 watts. Dividing 1505 by 70 gives me 21.50 watts. I’ll keep you updated on this. At midnight, I pushed it to 110th. At 12 am, it read 1025.78, then 1048.35 by 9 am. That’s 22.57 kW over 9 hours, about 2507 watts per hour. Doing the math gives me 2507 / 110 = 22.79 watts per TH. So that epic controller isn’t dropping power use, but it makes fan controls super easy. I do have 2 weak chips on this S19XP that drop out. Today I’m aiming for an up clock.

Saver more photos this is the 475 freq setting at 13 volts Note it gives false high on power in watts and one watts per th. I will get a reading on the meter 9:25 am. kwatts = 1049.39

Trust me, you don't want to overheat XPs. If you think its like the j Pro, think again...

Yeah I am in the 70c to 75c top heat area I may set shutdown to 82c not 85c My garage testing is over It looks like I can do 152th at 22 watts a th fans set to 90% and never go over 72c.  It is a decent board. tough call to say it is better than using braiins. the flat fee is nice but it is about 60 more than a plain controller. so if you earn 5 a day you save 10 cents a day over braiins takes 600 days to save the 60 dollars extra cost. not to mention the first 100 cost if you do the math that way it takes 1600 days.  now if everything takes off and you are making 30 a day then you save 60 cents a day and it is better than braiins If you are a home miner it is great for noise reduction. I do not think I want a lot of them for now. I am still wondering if I want to not buy a lot of xp's and go to s21's

Do the math against 97.5% of the Real Hashrate as shown in Braiins OS if you want to compare against "no dev fee" And don't forget pool fees... Its 0 in Braiins Pool. Ie: Shredd's 141T XP 1989W / (@104T*.975) 19.62 J/T

I like the physical controller its snappy better ram and cpu chip than bitmain’s controller. As for using your software vs the controller I will likely stick with yours over this. I have used epic vnish braiins + bitmain all with s19 s19j  s19a  s19xp  s19 pro  s19 pro j I prefer brains+ at this time. a

What was the best performance you managed to get on the S19xp? The base model is announced as 141T with 3031W. How long were you able to reduce your W?

basically the epic board on my 134th models does not increase efficiency at all. but it does lower fan speeds so if you are a home miner in you garage you can drop watts to about 2500 and do about 120th with fans at 60% So I recommend it if you want to make the miner quiet.

Based on your experience as a miner, do you think it is still possible to be a garage miner nowadays? I always had the "dream" of having mining equipment, but I never took this adventure. I think that every day that passes further reduces the possibility of having one and that it is minimally profitable.

be a garage miner and turn a profit. If you need heat and spend money on heat. 5 to 6 months every year If you have fairly cheap power. Maybe. My garage has 14 cent power 7 months a year and it can use the heat 5-6 months a year. The scrypt miner L7 turns a very small daily profit. If you conveniently leave out the purchase price of 3-5k. Yo could justify running it . I do this but that l7 paid itself off at my real mine. To go out and buy a machine for say 4000 that makes at my garage 3 bucks a day for only 7 months of the year. Is very hard to justify. Now if you get a cheap s19xp in a bit say 1800 and you have a need for heat yeah maybe .

Heating is not necessary. If you need it, it will only be 2-3 months a year. Advantage of living in Portugal.  In terms of energy costs, it will be around 15 cents. Anyway, it won't really be the best combinations for me to venture into having an ASIC. I don't know if you read my "diary" about an adventure I had in mining, using NiceHash. But, let me talk about this issue of energy: https://bitcointalk.org/index.php?topic=5336026.msg60372897#msg60372897 (Now I noticed that I didn't finish the story. I have to go take care of this.  )

Portugal nice weather.  Yeah you don't save much money for the heat it makes. Small garage mining can be done. 15 cents is high but not death. but at the moment you need wait for the right piece of gear at a good price. maybe the s19xp maybe  the s21 maybe an L7 maybe a E9 pro but all are very marginal at 15 cent power and all are 3000 usd and up. you kind of want a cheap one someone is dumping. you need to buy it you need a strong uptick in price of coin for the one you pick it is really difficult at 15 cents.

Exactly. But the best time to buy is also when the BTC value is low, as it allows you to find cheaper equipment. One day you discard some equipment, you might try to convince me to do business.

Just curious - has anyone had any experience with the epic boards on S19K Pro's? I am taking delivery of 15 K Pro's this week and was actually looking into the epic boards... I just don't like the concept of a dev fee so these stuck out. I'm not quite as concerned about maximum HR but rather maximum efficiency - I'm building out a facility in NY State where power + delivery comes in @ approximately 15 cents / kWh so getting every bit of efficiency out of these machines is definitely something I'd spend a bit extra on if it were a difference that I deemed worth dropping the coin on. Any/all feedback is very much appreciated. Cheers! -Jonathan

15 machines to run at 15 cents each really looks like a losing investment. the epic boards are nice but in your case I do not think they will help you.

Hey! First - I appreciate your feedback and Yes - you’re correct. It took a few weeks to get locked in to an industrial electricity rate (had to form an LLC and show articles of incorporation to get the power company to offer it and for whatever reason it took the department of treasury almost 60 days). I was looking to undervolt / under clock mostly to increase efficiency which the boards did actually work very well for. Tuning the 19K Pro’s to 106T with the ePIC’s voltage optimizer keeps them hashing at a stable 2350w (a few machines much more efficient - 1 or 2 much less lol). And I know lol - even at the. 11.2 cents / kWh the numbers are a bit iffy - but I’ve got some S21’s that I’ve started to take delivery of so running these alongside those until I retire them seems pretty decent comparatively looking at other machines at that price point. With regard to efficiency, the UMC’s definitely helped me get to where I wanted to be. So I guess what I want to add here is this - for anyone considering them, I can’t speak to how far you can push the chips/boards for maximum HR, but if you’re looking to max out the efficiency while maintaining an impactful hashrate, the boards are a solid choice. And yeah I paid for the machines in late September so I paid closer to $1500 USD/S19k Pro. Anticipating the machines will follow the market price of BTC as they have in past cycles.